WEBVTT 00:00:00.000 --> 00:00:01.988 align:middle line:90% [SQUEAKING] 00:00:01.988 --> 00:00:03.476 align:middle line:90% [RUSTLING] 00:00:03.476 --> 00:00:05.956 align:middle line:90% [CLICKING] 00:00:05.956 --> 00:00:17.000 align:middle line:90% 00:00:17.000 --> 00:00:18.693 align:middle line:84% JUEJUN HU: OK, so in the last lecture, 00:00:18.693 --> 00:00:20.360 align:middle line:84% we talked about essentially the material 00:00:20.360 --> 00:00:22.980 align:middle line:84% that you use to make the breakable screens. 00:00:22.980 --> 00:00:26.060 align:middle line:84% [LAUGHS] So essentially, Gorilla glass, right? 00:00:26.060 --> 00:00:28.840 align:middle line:84% But there are also a lot of other competing glass materials 00:00:28.840 --> 00:00:31.880 align:middle line:84% from other companies essentially exhibiting similar properties. 00:00:31.880 --> 00:00:34.480 align:middle line:84% The whole idea is that you do this ion exchange to get 00:00:34.480 --> 00:00:37.200 align:middle line:84% a very large surface compressive stress, which 00:00:37.200 --> 00:00:40.960 align:middle line:84% helps you to essentially prevent the crack propagation. 00:00:40.960 --> 00:00:43.440 align:middle line:84% Now, I just want to start by mentioning 00:00:43.440 --> 00:00:46.420 align:middle line:84% that this ion exchange process does a couple of things. 00:00:46.420 --> 00:00:50.462 align:middle line:90% 00:00:50.462 --> 00:00:51.920 align:middle line:84% OK, what I'm showing you here, this 00:00:51.920 --> 00:00:56.120 align:middle line:84% is the Weibull plot of essentially glasses 00:00:56.120 --> 00:00:58.738 align:middle line:84% with and without the ion exchange process. 00:00:58.738 --> 00:01:00.280 align:middle line:84% So if you look at Weibull plot, there 00:01:00.280 --> 00:01:01.840 align:middle line:84% are two things you can tell here. 00:01:01.840 --> 00:01:04.720 align:middle line:84% One is that the curve gets shifted 00:01:04.720 --> 00:01:09.480 align:middle line:84% towards the right side for the chemically strengthened glass, 00:01:09.480 --> 00:01:14.880 align:middle line:84% which indicates that you have now a higher strength. 00:01:14.880 --> 00:01:16.760 align:middle line:84% It can withstand essentially higher stress 00:01:16.760 --> 00:01:17.660 align:middle line:90% without fracturing. 00:01:17.660 --> 00:01:18.880 align:middle line:90% Makes sense, right? 00:01:18.880 --> 00:01:22.840 align:middle line:84% Now the second thing is you notice the slope of the line. 00:01:22.840 --> 00:01:25.800 align:middle line:90% The slope of the line increases. 00:01:25.800 --> 00:01:28.040 align:middle line:84% If you remember, the slope essentially 00:01:28.040 --> 00:01:30.560 align:middle line:84% corresponds to the Weibull modulus. 00:01:30.560 --> 00:01:34.640 align:middle line:84% And the larger the modulus, the more consistent 00:01:34.640 --> 00:01:35.820 align:middle line:90% the glass material is. 00:01:35.820 --> 00:01:37.737 align:middle line:84% Essentially, it means that the glass material, 00:01:37.737 --> 00:01:39.920 align:middle line:84% they have a much narrower distribution of defects. 00:01:39.920 --> 00:01:42.100 align:middle line:84% So that's also what's shown here. 00:01:42.100 --> 00:01:44.540 align:middle line:84% Essentially, with the ion exchange process, 00:01:44.540 --> 00:01:48.200 align:middle line:84% it essentially will give you a more reproducible or more 00:01:48.200 --> 00:01:51.955 align:middle line:84% consistent product finish compared to untreated. 00:01:51.955 --> 00:01:53.580 align:middle line:84% If it's just a regular soda lime glass, 00:01:53.580 --> 00:01:55.640 align:middle line:84% it's a much wider distribution of defects 00:01:55.640 --> 00:01:57.700 align:middle line:90% in terms of size and severity. 00:01:57.700 --> 00:02:01.600 align:middle line:90% 00:02:01.600 --> 00:02:05.260 align:middle line:84% So that's why people like to use these strengthened glasses, 00:02:05.260 --> 00:02:08.620 align:middle line:84% which is now a standard on most consumer electronic products. 00:02:08.620 --> 00:02:10.259 align:middle line:84% They are the cover glass materials. 00:02:10.259 --> 00:02:14.160 align:middle line:84% Now, next thing I want to do, I want to actually draw 00:02:14.160 --> 00:02:15.720 align:middle line:90% a little comparison. 00:02:15.720 --> 00:02:20.585 align:middle line:84% Glasses do not come without competitors. 00:02:20.585 --> 00:02:21.960 align:middle line:84% For a long time, people have been 00:02:21.960 --> 00:02:24.220 align:middle line:84% talking about alternative carbon materials. 00:02:24.220 --> 00:02:27.400 align:middle line:84% And sapphire is the one that really stands out. 00:02:27.400 --> 00:02:30.180 align:middle line:84% In fact, if you go back about 10 years, 00:02:30.180 --> 00:02:32.240 align:middle line:84% there was a lot of discussions about Apple 00:02:32.240 --> 00:02:39.340 align:middle line:84% is going to switch to sapphire in the iPhone cover materials. 00:02:39.340 --> 00:02:43.440 align:middle line:84% If you look back in like 2014, 2015-ish kind of time frame, 00:02:43.440 --> 00:02:47.760 align:middle line:84% Apple actually invested in a company, this one, GT ASF, 00:02:47.760 --> 00:02:49.220 align:middle line:90% GT Advanced Materials. 00:02:49.220 --> 00:02:51.500 align:middle line:84% The link is actually no longer working, 00:02:51.500 --> 00:02:54.440 align:middle line:84% but I decided to leave it there because the company has already 00:02:54.440 --> 00:02:56.240 align:middle line:90% been acquired. 00:02:56.240 --> 00:02:58.950 align:middle line:84% So Apple actually invested in that company several hundred 00:02:58.950 --> 00:03:02.590 align:middle line:84% of millions of dollars just to essentially crank up 00:03:02.590 --> 00:03:06.050 align:middle line:84% the manufacturing capability for sapphire screens. 00:03:06.050 --> 00:03:07.955 align:middle line:84% It unfortunately did not really work out. 00:03:07.955 --> 00:03:09.830 align:middle line:84% If you look back the news, there was actually 00:03:09.830 --> 00:03:13.910 align:middle line:84% a lot of drama in that whole deal there. 00:03:13.910 --> 00:03:16.990 align:middle line:84% The company eventually filed for bankruptcy, 00:03:16.990 --> 00:03:19.362 align:middle line:84% but it doesn't mean that sapphire is going away. 00:03:19.362 --> 00:03:21.570 align:middle line:84% Apple still uses sapphire for some of their products, 00:03:21.570 --> 00:03:23.510 align:middle line:90% especially Apple Watch. 00:03:23.510 --> 00:03:26.730 align:middle line:84% So any of you have the Apple Watch? 00:03:26.730 --> 00:03:27.230 align:middle line:90% No? 00:03:27.230 --> 00:03:28.353 align:middle line:90% Just one, OK. 00:03:28.353 --> 00:03:29.770 align:middle line:84% I don't know which model you have. 00:03:29.770 --> 00:03:31.430 align:middle line:90% AUDIENCE: It's old. 00:03:31.430 --> 00:03:32.180 align:middle line:90% JUEJUN HU: OK. 00:03:32.180 --> 00:03:35.030 align:middle line:84% [LAUGHS] They have essentially two product lines. 00:03:35.030 --> 00:03:36.810 align:middle line:90% One is the lower-end model. 00:03:36.810 --> 00:03:38.390 align:middle line:90% They use aluminum casing. 00:03:38.390 --> 00:03:41.310 align:middle line:84% On that one, they still use the strengthened glass. 00:03:41.310 --> 00:03:42.770 align:middle line:84% They also have a higher-end model 00:03:42.770 --> 00:03:45.190 align:middle line:84% that uses either stainless steel or titanium. 00:03:45.190 --> 00:03:46.930 align:middle line:90% That one-- sapphire screen. 00:03:46.930 --> 00:03:50.330 align:middle line:84% I don't know which one you have, but you can try to scratch it. 00:03:50.330 --> 00:03:50.830 align:middle line:90% No. 00:03:50.830 --> 00:03:53.370 align:middle line:84% [LAUGHS] If you scratch it and it leaves a scratch, then 00:03:53.370 --> 00:03:53.870 align:middle line:90% it's glass. 00:03:53.870 --> 00:03:54.530 align:middle line:90% No, don't try it. 00:03:54.530 --> 00:03:55.550 align:middle line:84% AUDIENCE: It's pretty scratched, so I'm 00:03:55.550 --> 00:03:56.850 align:middle line:90% going to say is not sapphire. 00:03:56.850 --> 00:03:58.510 align:middle line:84% JUEJUN HU: It's probably glass, then. 00:03:58.510 --> 00:03:59.170 align:middle line:90% Don't try it. 00:03:59.170 --> 00:04:01.670 align:middle line:84% That's the best way to create surface defect that eventually 00:04:01.670 --> 00:04:02.993 align:middle line:90% lead to catastrophic failure. 00:04:02.993 --> 00:04:03.910 align:middle line:90% AUDIENCE: It is rough. 00:04:03.910 --> 00:04:04.870 align:middle line:90% It might happen. 00:04:04.870 --> 00:04:06.310 align:middle line:90% [LAUGHS] 00:04:06.310 --> 00:04:07.350 align:middle line:90% JUEJUN HU: Probably. 00:04:07.350 --> 00:04:09.910 align:middle line:90% Do your fractography. 00:04:09.910 --> 00:04:11.230 align:middle line:90% There you go. 00:04:11.230 --> 00:04:13.365 align:middle line:84% OK, so now what I want to do, I want 00:04:13.365 --> 00:04:15.990 align:middle line:84% to actually draw a little bit of a comparison between these two 00:04:15.990 --> 00:04:18.110 align:middle line:90% commonly used cover materials. 00:04:18.110 --> 00:04:19.709 align:middle line:84% And in the end, what I tried to show 00:04:19.709 --> 00:04:22.050 align:middle line:84% you here is that Apple actually made some rational choices. 00:04:22.050 --> 00:04:24.633 align:middle line:84% There's a good reason why they use this specifically for Apple 00:04:24.633 --> 00:04:27.390 align:middle line:90% Watch in the higher-end models. 00:04:27.390 --> 00:04:29.390 align:middle line:84% So starting by looking at optical properties, 00:04:29.390 --> 00:04:32.550 align:middle line:84% refractive index, sapphire has a slightly higher refractive 00:04:32.550 --> 00:04:33.050 align:middle line:90% index. 00:04:33.050 --> 00:04:33.990 align:middle line:90% What does that mean? 00:04:33.990 --> 00:04:36.590 align:middle line:84% Refractive index actually impacts optical reflection. 00:04:36.590 --> 00:04:39.190 align:middle line:84% The higher refractive index, the more reflection you get. 00:04:39.190 --> 00:04:43.590 align:middle line:84% So for glass, typical, normal, surface-normal refraction 00:04:43.590 --> 00:04:46.670 align:middle line:84% is about 4% based on the refractive index value. 00:04:46.670 --> 00:04:51.350 align:middle line:84% For sapphire, it's about 7.5%, 8%. 00:04:51.350 --> 00:04:53.710 align:middle line:90% So it's not a big difference. 00:04:53.710 --> 00:04:55.230 align:middle line:84% And you can potentially mitigate it 00:04:55.230 --> 00:04:58.070 align:middle line:84% with essentially anti-reflection coatings. 00:04:58.070 --> 00:05:02.510 align:middle line:84% But glass does have a slightly lower surface reflectance 00:05:02.510 --> 00:05:03.690 align:middle line:90% compared to sapphire. 00:05:03.690 --> 00:05:07.630 align:middle line:90% 00:05:07.630 --> 00:05:10.270 align:middle line:90% Next part is density. 00:05:10.270 --> 00:05:13.130 align:middle line:90% So sapphire has higher density. 00:05:13.130 --> 00:05:14.670 align:middle line:84% Now for a cover material, you want 00:05:14.670 --> 00:05:18.510 align:middle line:84% it to have a smaller mass because everyone wants lighter 00:05:18.510 --> 00:05:19.930 align:middle line:90% consumer electronic devices. 00:05:19.930 --> 00:05:22.510 align:middle line:84% Again, it doesn't really make a big difference 00:05:22.510 --> 00:05:25.510 align:middle line:84% for small consumer electronic devices like Apple Watch. 00:05:25.510 --> 00:05:29.450 align:middle line:84% For larger devices like iPad, that weight actually adds up. 00:05:29.450 --> 00:05:33.270 align:middle line:90% 00:05:33.270 --> 00:05:36.390 align:middle line:84% All right, next one, this is one of the most important metric 00:05:36.390 --> 00:05:38.430 align:middle line:90% here-- fracture toughness. 00:05:38.430 --> 00:05:40.330 align:middle line:84% So sapphire has a fracture toughness 00:05:40.330 --> 00:05:43.390 align:middle line:84% that's more than 2 times higher or 3 times higher 00:05:43.390 --> 00:05:45.470 align:middle line:90% compared to glass materials. 00:05:45.470 --> 00:05:48.550 align:middle line:84% And that's why if you actually read news report, 00:05:48.550 --> 00:05:50.030 align:middle line:84% some people were actually claiming 00:05:50.030 --> 00:05:55.370 align:middle line:84% sapphire is 3 times stronger compared to strengthened glass. 00:05:55.370 --> 00:05:57.502 align:middle line:84% We should give credit to authors of those articles 00:05:57.502 --> 00:05:59.210 align:middle line:84% because they actually did their homework. 00:05:59.210 --> 00:06:01.252 align:middle line:84% They actually looked up their fracture toughness, 00:06:01.252 --> 00:06:04.830 align:middle line:84% and that's where they draw their quantitative conclusion. 00:06:04.830 --> 00:06:07.370 align:middle line:84% Now, however, what we have mentioned in the last lecture, 00:06:07.370 --> 00:06:09.750 align:middle line:84% if you recall, the glass, the fracture toughness 00:06:09.750 --> 00:06:15.070 align:middle line:84% was actually measured using ion-strengthened glass samples. 00:06:15.070 --> 00:06:16.950 align:middle line:84% It did not account for the fact that if you 00:06:16.950 --> 00:06:18.790 align:middle line:84% have compressive stress, it actually 00:06:18.790 --> 00:06:20.750 align:middle line:84% helps to prevent crack propagation. 00:06:20.750 --> 00:06:22.430 align:middle line:84% So the apparent fracture toughness 00:06:22.430 --> 00:06:24.110 align:middle line:84% or the apparent ability for glass 00:06:24.110 --> 00:06:26.750 align:middle line:84% to resist crack propagation is higher than what's actually 00:06:26.750 --> 00:06:28.470 align:middle line:90% specified by that number there. 00:06:28.470 --> 00:06:31.970 align:middle line:84% So the 3 times higher fracture toughness or 3 times stronger, 00:06:31.970 --> 00:06:36.030 align:middle line:84% that's not a totally fair comparison. 00:06:36.030 --> 00:06:38.150 align:middle line:84% Now, in fact, we can compare this kind 00:06:38.150 --> 00:06:40.630 align:middle line:84% of availability of glass to resist fracture 00:06:40.630 --> 00:06:42.310 align:middle line:84% by looking at their strength here. 00:06:42.310 --> 00:06:44.105 align:middle line:84% So you can look at tensile strength. 00:06:44.105 --> 00:06:46.730 align:middle line:84% Unfortunately, I didn't find the number for strengthened glass. 00:06:46.730 --> 00:06:49.470 align:middle line:84% But flexural strength is what's more relevant. 00:06:49.470 --> 00:06:53.980 align:middle line:84% So flexural stress is basically the maximum surface stress 00:06:53.980 --> 00:06:57.938 align:middle line:84% the material can withstand before it breaks during bending. 00:06:57.938 --> 00:06:59.980 align:middle line:84% So you can think of it as the ability of material 00:06:59.980 --> 00:07:01.820 align:middle line:90% to resist bending. 00:07:01.820 --> 00:07:03.380 align:middle line:90% Don't try it. 00:07:03.380 --> 00:07:04.880 align:middle line:84% You're not going to break it anyway. 00:07:04.880 --> 00:07:08.460 align:middle line:84% But if you compare them, sapphire and strengthened glass 00:07:08.460 --> 00:07:11.780 align:middle line:84% actually has quite comparable flexural strength. 00:07:11.780 --> 00:07:14.140 align:middle line:84% So in that regard, their performance 00:07:14.140 --> 00:07:16.140 align:middle line:84% is similar in this case, the ability 00:07:16.140 --> 00:07:22.460 align:middle line:84% to essentially withstand any kind of deformation. 00:07:22.460 --> 00:07:25.220 align:middle line:84% Now, next one, the modulus, this is 00:07:25.220 --> 00:07:27.540 align:middle line:84% where you can make quite a bit of difference. 00:07:27.540 --> 00:07:30.780 align:middle line:84% Sapphire is a very stiff material. 00:07:30.780 --> 00:07:35.060 align:middle line:84% It has a Young's modulus that's 5 times higher compared 00:07:35.060 --> 00:07:36.120 align:middle line:90% to typical glasses. 00:07:36.120 --> 00:07:39.380 align:middle line:90% 00:07:39.380 --> 00:07:41.000 align:middle line:90% So what's the consequence? 00:07:41.000 --> 00:07:43.500 align:middle line:84% If you're trying to just look at the strength of essentially 00:07:43.500 --> 00:07:46.190 align:middle line:84% the screen, stand-alone screen, whether sapphire or glass, 00:07:46.190 --> 00:07:47.940 align:middle line:84% they are more or less similar because they 00:07:47.940 --> 00:07:49.820 align:middle line:90% have similar flexural strength. 00:07:49.820 --> 00:07:54.060 align:middle line:84% However, when you put them in an electronic device, 00:07:54.060 --> 00:07:56.940 align:middle line:84% there are other materials that actually goes around the screen. 00:07:56.940 --> 00:07:58.780 align:middle line:90% So when you actually drop it-- 00:07:58.780 --> 00:08:00.180 align:middle line:90% I'm not going to try it-- 00:08:00.180 --> 00:08:02.420 align:middle line:84% if you actually drop it, then what actually deforms 00:08:02.420 --> 00:08:05.380 align:middle line:84% is not just your screen, but it's essentially the whole form. 00:08:05.380 --> 00:08:07.540 align:middle line:84% So now imagine if you have your screen 00:08:07.540 --> 00:08:08.842 align:middle line:90% as the most stiff component. 00:08:08.842 --> 00:08:11.300 align:middle line:84% For example, if you have a sapphire, then when you actually 00:08:11.300 --> 00:08:14.780 align:middle line:84% drop it, most of the stress is going 00:08:14.780 --> 00:08:16.500 align:middle line:84% to be exerted on your screen because this 00:08:16.500 --> 00:08:18.458 align:middle line:84% is a stiff material, stiffest the material that 00:08:18.458 --> 00:08:19.860 align:middle line:90% has the least deformation. 00:08:19.860 --> 00:08:21.440 align:middle line:90% It will resist deformation most. 00:08:21.440 --> 00:08:24.940 align:middle line:90% 00:08:24.940 --> 00:08:27.500 align:middle line:84% And if you think about the Young's modulus 00:08:27.500 --> 00:08:28.880 align:middle line:84% of other materials like aluminum, 00:08:28.880 --> 00:08:31.420 align:middle line:84% aluminum has a Young's modulus of about 70 gigapascal. 00:08:31.420 --> 00:08:34.360 align:middle line:84% So that's comparable to glass but much lower than sapphire. 00:08:34.360 --> 00:08:37.460 align:middle line:84% So it means that essentially sapphire on the aluminum 00:08:37.460 --> 00:08:40.220 align:middle line:84% casing form is going to withstand a lot larger stress 00:08:40.220 --> 00:08:42.380 align:middle line:90% compared to glass. 00:08:42.380 --> 00:08:43.820 align:middle line:90% Now what about stainless steel? 00:08:43.820 --> 00:08:46.140 align:middle line:84% Stainless steel has a Young's modulus higher than 200 00:08:46.140 --> 00:08:48.220 align:middle line:90% gigapascal. 00:08:48.220 --> 00:08:51.380 align:middle line:84% So the mismatch is a lot smaller. 00:08:51.380 --> 00:08:53.140 align:middle line:84% So you see that actually makes sense 00:08:53.140 --> 00:08:56.340 align:middle line:84% that for stainless-steel-cased Apple Watch, 00:08:56.340 --> 00:08:58.340 align:middle line:84% people will actually use a sapphire screen 00:08:58.340 --> 00:09:02.500 align:middle line:84% because the modulus mismatch is a lot less compared to casing 00:09:02.500 --> 00:09:03.820 align:middle line:90% and the screen there. 00:09:03.820 --> 00:09:05.820 align:middle line:90% In fact, it's not just Apple. 00:09:05.820 --> 00:09:09.060 align:middle line:84% If you look at other premium electronic manufacturers 00:09:09.060 --> 00:09:11.280 align:middle line:84% like Samsung, they make similar choices. 00:09:11.280 --> 00:09:14.820 align:middle line:84% They have sapphire screens on their Galaxy Watch 00:09:14.820 --> 00:09:16.340 align:middle line:84% or whatever, the high-end models. 00:09:16.340 --> 00:09:19.100 align:middle line:84% So again, out of that reasoning, it 00:09:19.100 --> 00:09:21.660 align:middle line:84% makes sense that they use it for casing 00:09:21.660 --> 00:09:22.900 align:middle line:90% with high Young's modulus. 00:09:22.900 --> 00:09:27.820 align:middle line:84% Now last one, this is where sapphire really shines-- 00:09:27.820 --> 00:09:29.660 align:middle line:90% hardness. 00:09:29.660 --> 00:09:32.020 align:middle line:84% Sapphire is one of the hardest materials, 00:09:32.020 --> 00:09:34.820 align:middle line:84% I think, basically, only second to diamond. 00:09:34.820 --> 00:09:39.180 align:middle line:84% And this, of course, shows up as a very high Vickers hardness. 00:09:39.180 --> 00:09:40.900 align:middle line:84% Harness essentially means that it's 00:09:40.900 --> 00:09:42.880 align:middle line:90% a lot more scratch resistant. 00:09:42.880 --> 00:09:45.500 align:middle line:90% 00:09:45.500 --> 00:09:48.940 align:middle line:84% And surface scratches is usually the defect 00:09:48.940 --> 00:09:52.140 align:middle line:84% that eventually leads to breaking of your phone screen 00:09:52.140 --> 00:09:54.340 align:middle line:90% or your whatever, iPad screen. 00:09:54.340 --> 00:09:58.160 align:middle line:84% So the risk, if you put, for example, 00:09:58.160 --> 00:10:04.240 align:middle line:84% your phone in your wallet, in your pocket, 00:10:04.240 --> 00:10:07.200 align:middle line:84% along with other stiff materials like your keys, 00:10:07.200 --> 00:10:09.200 align:middle line:84% or I don't know where you put it in your pocket, 00:10:09.200 --> 00:10:13.660 align:middle line:84% but the risk of causing surface scratch is higher for glass 00:10:13.660 --> 00:10:14.760 align:middle line:90% compared to sapphire. 00:10:14.760 --> 00:10:18.600 align:middle line:90% 00:10:18.600 --> 00:10:20.100 align:middle line:84% So despite the fact that people have 00:10:20.100 --> 00:10:22.420 align:middle line:84% made major strides over the past few years 00:10:22.420 --> 00:10:24.800 align:middle line:84% in terms of the strength of screens, 00:10:24.800 --> 00:10:26.420 align:middle line:84% my recommendation is still, try not 00:10:26.420 --> 00:10:29.180 align:middle line:84% to put it with something that's sharp and stiff. 00:10:29.180 --> 00:10:30.860 align:middle line:84% It may not break it immediately, but it 00:10:30.860 --> 00:10:32.700 align:middle line:90% may cause more surface cracks. 00:10:32.700 --> 00:10:35.540 align:middle line:84% And in those cracks eventually is where 00:10:35.540 --> 00:10:37.500 align:middle line:90% the defects will nucleate. 00:10:37.500 --> 00:10:39.740 align:middle line:84% And that's where the Griffith's crack comes along. 00:10:39.740 --> 00:10:41.782 align:middle line:84% You will get stress concentration and everything. 00:10:41.782 --> 00:10:44.800 align:middle line:84% So again, that's where sapphire really wins. 00:10:44.800 --> 00:10:48.930 align:middle line:84% AUDIENCE: Is that why screen protectors are a thing? 00:10:48.930 --> 00:10:49.570 align:middle line:90% JUEJUN HU: Yes. 00:10:49.570 --> 00:10:52.250 align:middle line:84% AUDIENCE: Because it just prevents the scratches. 00:10:52.250 --> 00:10:53.570 align:middle line:90% JUEJUN HU: Not really. 00:10:53.570 --> 00:10:55.130 align:middle line:84% The screen protectors essentially 00:10:55.130 --> 00:10:57.810 align:middle line:84% is like, they take the scratches on the screen protectors 00:10:57.810 --> 00:10:59.810 align:middle line:84% so the underlying glass doesn't get damaged. 00:10:59.810 --> 00:11:03.990 align:middle line:84% So yes, but the downside is that you reduce light transmission. 00:11:03.990 --> 00:11:06.573 align:middle line:90% 00:11:06.573 --> 00:11:07.990 align:middle line:84% That's the biggest thing, I think. 00:11:07.990 --> 00:11:10.410 align:middle line:90% So it's really up to you. 00:11:10.410 --> 00:11:12.170 align:middle line:84% I don't use any screen protector because I 00:11:12.170 --> 00:11:13.930 align:middle line:84% feel like people essentially make so much effort trying 00:11:13.930 --> 00:11:15.170 align:middle line:84% to make it more scratch resistant 00:11:15.170 --> 00:11:16.462 align:middle line:90% and make it highly transparent. 00:11:16.462 --> 00:11:19.210 align:middle line:84% They put all kinds of coatings to increase light transmittance, 00:11:19.210 --> 00:11:20.668 align:middle line:84% and you just put a screen protector 00:11:20.668 --> 00:11:23.050 align:middle line:84% on top of that to negate all those efforts. 00:11:23.050 --> 00:11:25.510 align:middle line:84% So out of respect for the glass engineers, 00:11:25.510 --> 00:11:27.850 align:middle line:84% I decided not to put the screen protector. 00:11:27.850 --> 00:11:30.330 align:middle line:84% AUDIENCE: Why do you want to match your casing 00:11:30.330 --> 00:11:32.510 align:middle line:90% and screen's Young's moduli? 00:11:32.510 --> 00:11:34.650 align:middle line:84% JUEJUN HU: So basically, what happens 00:11:34.650 --> 00:11:37.810 align:middle line:84% when you actually drop it, the whole thing deforms. 00:11:37.810 --> 00:11:40.890 align:middle line:84% If everything deforms more or less with the same extent, 00:11:40.890 --> 00:11:42.670 align:middle line:84% then you have the same level of strain. 00:11:42.670 --> 00:11:43.387 align:middle line:90% AUDIENCE: Oh, OK. 00:11:43.387 --> 00:11:45.470 align:middle line:84% JUEJUN HU: But if one thing is particularly stiff, 00:11:45.470 --> 00:11:47.430 align:middle line:84% everything else is soft in the extreme case. 00:11:47.430 --> 00:11:49.670 align:middle line:84% Essentially, all the soft parts can easily deform, 00:11:49.670 --> 00:11:54.410 align:middle line:84% but then the stiff part actually withstand the much larger stress 00:11:54.410 --> 00:11:55.948 align:middle line:90% compared to anything else. 00:11:55.948 --> 00:11:56.490 align:middle line:90% AUDIENCE: OK. 00:11:56.490 --> 00:11:57.890 align:middle line:84% JUEJUN HU: So you actually don't want that. 00:11:57.890 --> 00:12:00.730 align:middle line:84% You want everything to deform more or less to the same extent. 00:12:00.730 --> 00:12:02.330 align:middle line:84% That's why ideally everything should 00:12:02.330 --> 00:12:04.735 align:middle line:84% have more or less the same mechanical stiffness. 00:12:04.735 --> 00:12:06.110 align:middle line:84% AUDIENCE: Yeah, that makes sense. 00:12:06.110 --> 00:12:06.690 align:middle line:90% OK, cool. 00:12:06.690 --> 00:12:07.357 align:middle line:90% JUEJUN HU: Yeah. 00:12:07.357 --> 00:12:09.930 align:middle line:84% So that's a good reason why people should 00:12:09.930 --> 00:12:12.810 align:middle line:90% use one material or the other. 00:12:12.810 --> 00:12:14.530 align:middle line:90% Make sense? 00:12:14.530 --> 00:12:17.770 align:middle line:84% Now, outside these apparent quantitative material 00:12:17.770 --> 00:12:21.650 align:middle line:84% properties, obviously, there's also a perception difference 00:12:21.650 --> 00:12:24.510 align:middle line:90% for these two materials. 00:12:24.510 --> 00:12:29.490 align:middle line:84% Sapphire, to the years of known material scientists, 00:12:29.490 --> 00:12:31.970 align:middle line:90% it's a precious gemstone. 00:12:31.970 --> 00:12:36.073 align:middle line:84% Now, to us it's basically nothing more than Al2O3. 00:12:36.073 --> 00:12:38.490 align:middle line:84% I'm actually surprised people are not calling alumina, why 00:12:38.490 --> 00:12:39.830 align:middle line:90% they call it a sapphire screen. 00:12:39.830 --> 00:12:42.210 align:middle line:84% They should call it alumina screen. 00:12:42.210 --> 00:12:46.510 align:middle line:84% Anyway, so sapphire does give people a feeling. 00:12:46.510 --> 00:12:49.670 align:middle line:84% In fact, people call it a luxury cover material. 00:12:49.670 --> 00:12:51.797 align:middle line:84% Nobody calls glass a luxury cover material, 00:12:51.797 --> 00:12:53.630 align:middle line:84% even though they last for billions of years. 00:12:53.630 --> 00:12:57.088 align:middle line:90% 00:12:57.088 --> 00:12:59.130 align:middle line:84% So anyway, that's also part of the reason they're 00:12:59.130 --> 00:13:00.930 align:middle line:90% used in the high-end models. 00:13:00.930 --> 00:13:04.890 align:middle line:84% Also, sapphire is indeed more expensive than glass. 00:13:04.890 --> 00:13:07.690 align:middle line:84% The cost of sapphire substrates or windows 00:13:07.690 --> 00:13:11.690 align:middle line:84% has gone down very significantly over the past couple of decades, 00:13:11.690 --> 00:13:13.890 align:middle line:84% not because of the display industry, 00:13:13.890 --> 00:13:17.930 align:middle line:84% in fact, because of the lighting industry, LEDs. 00:13:17.930 --> 00:13:20.570 align:middle line:84% Blue LEDs, gallium nitride, is often 00:13:20.570 --> 00:13:22.810 align:middle line:90% grown on sapphire substrate. 00:13:22.810 --> 00:13:26.030 align:middle line:84% So because of the economy of scale coming from LED industry, 00:13:26.030 --> 00:13:28.027 align:middle line:84% that's why sapphire gets a lot cheaper 00:13:28.027 --> 00:13:29.110 align:middle line:90% over the past two decades. 00:13:29.110 --> 00:13:31.050 align:middle line:84% And that's what makes it viable to actually make it even out 00:13:31.050 --> 00:13:31.970 align:middle line:90% of screens. 00:13:31.970 --> 00:13:34.670 align:middle line:84% But still, anyway, it's still more expensive than glass. 00:13:34.670 --> 00:13:36.930 align:middle line:84% That's why it also makes sense Apple puts it only 00:13:36.930 --> 00:13:38.630 align:middle line:84% in essentially the high-end models. 00:13:38.630 --> 00:13:43.410 align:middle line:90% 00:13:43.410 --> 00:13:47.472 align:middle line:84% OK, so that's the comparison here. 00:13:47.472 --> 00:13:48.930 align:middle line:84% But now what I also want to mention 00:13:48.930 --> 00:13:52.430 align:middle line:84% is that, again, people have been working on glasses for decades, 00:13:52.430 --> 00:13:53.750 align:middle line:90% these strengthened glasses. 00:13:53.750 --> 00:13:55.490 align:middle line:84% They've come up with new ways to further improve 00:13:55.490 --> 00:13:57.115 align:middle line:84% the glass properties beyond what I have 00:13:57.115 --> 00:13:59.250 align:middle line:90% listed in the previous table. 00:13:59.250 --> 00:14:03.610 align:middle line:84% So one example here, Corning has this line 00:14:03.610 --> 00:14:07.330 align:middle line:84% of products they call scratch-resistant glasses. 00:14:07.330 --> 00:14:10.970 align:middle line:84% The first generation is what they label as SR+. 00:14:10.970 --> 00:14:13.690 align:middle line:84% The more recent one they have a DX+. 00:14:13.690 --> 00:14:16.650 align:middle line:84% Unfortunately, they did not give much information about what DX+ 00:14:16.650 --> 00:14:17.150 align:middle line:90% is. 00:14:17.150 --> 00:14:18.810 align:middle line:90% We know it's derived from SR+. 00:14:18.810 --> 00:14:21.110 align:middle line:84% So that's why I'm using SR+ still as an example, 00:14:21.110 --> 00:14:23.730 align:middle line:84% even though it's a somewhat outdated product. 00:14:23.730 --> 00:14:25.890 align:middle line:84% So what they are showing here, essentially, this 00:14:25.890 --> 00:14:29.030 align:middle line:84% is a little demonstration experiment. 00:14:29.030 --> 00:14:32.075 align:middle line:84% If you use the regular Gorilla Glass, you start to scratch it, 00:14:32.075 --> 00:14:33.450 align:middle line:84% this is essentially the scratches 00:14:33.450 --> 00:14:34.930 align:middle line:84% that happens using different forces you 00:14:34.930 --> 00:14:36.090 align:middle line:90% apply on the glass screen. 00:14:36.090 --> 00:14:38.780 align:middle line:84% You see a very large scratch happening here. 00:14:38.780 --> 00:14:43.840 align:middle line:84% If you use either the Gorilla Glass SR+ or this one, ALCM. 00:14:43.840 --> 00:14:47.280 align:middle line:84% It stands for Alternative Luxury Cover 00:14:47.280 --> 00:14:49.900 align:middle line:90% Material, basically sapphire. 00:14:49.900 --> 00:14:53.000 align:middle line:84% You see that you don't get any scratches on these two 00:14:53.000 --> 00:14:55.040 align:middle line:90% materials. 00:14:55.040 --> 00:14:58.560 align:middle line:84% And so they did not give much information 00:14:58.560 --> 00:15:02.560 align:middle line:84% about why this glass material is more scratch resistant compared 00:15:02.560 --> 00:15:05.080 align:middle line:90% to the untreated Gorilla Glass. 00:15:05.080 --> 00:15:09.120 align:middle line:84% But as an MIT engineer, we like to dig a little bit deeper, 00:15:09.120 --> 00:15:12.560 align:middle line:84% so I actually look at their product spec sheet. 00:15:12.560 --> 00:15:15.080 align:middle line:84% I found in one corner, they actually 00:15:15.080 --> 00:15:18.600 align:middle line:84% give this transmission spectrum here. 00:15:18.600 --> 00:15:20.660 align:middle line:84% The main purpose they give that diagram 00:15:20.660 --> 00:15:22.600 align:middle line:84% is to show that the glass material actually 00:15:22.600 --> 00:15:25.400 align:middle line:84% has a high optical transmission compared to sapphire, 00:15:25.400 --> 00:15:27.600 align:middle line:84% basically because of that lower refractive index. 00:15:27.600 --> 00:15:31.940 align:middle line:84% But what I also noticed is that unlike untreated Gorilla Glass, 00:15:31.940 --> 00:15:34.200 align:middle line:84% which has a more or less flat spectrum, 00:15:34.200 --> 00:15:38.360 align:middle line:84% you see these fringes or fluctuations coming up 00:15:38.360 --> 00:15:40.280 align:middle line:84% on the transmission spectrum there. 00:15:40.280 --> 00:15:43.880 align:middle line:84% So my speculation is that this is not just 00:15:43.880 --> 00:15:45.720 align:middle line:90% one uniform glass material. 00:15:45.720 --> 00:15:48.480 align:middle line:84% What happens is that there can be one layer or maybe 00:15:48.480 --> 00:15:51.000 align:middle line:84% multiple layers of some surface coating 00:15:51.000 --> 00:15:52.860 align:middle line:84% so that when the light comes through, 00:15:52.860 --> 00:15:55.240 align:middle line:84% it can get reflected either on the top surface 00:15:55.240 --> 00:16:00.000 align:middle line:84% or it can actually get reflected at the glass coating interface. 00:16:00.000 --> 00:16:03.060 align:middle line:84% Now, depending on the wavelength and thickness of the film, 00:16:03.060 --> 00:16:04.680 align:middle line:84% these two reflected light waves, they 00:16:04.680 --> 00:16:07.960 align:middle line:84% can interfere either constructively or destructively, 00:16:07.960 --> 00:16:09.760 align:middle line:84% which then lead to an intensity fluctuation 00:16:09.760 --> 00:16:11.040 align:middle line:90% as a function of wavelength. 00:16:11.040 --> 00:16:15.220 align:middle line:84% So I actually wrote the Matlab code 00:16:15.220 --> 00:16:18.040 align:middle line:84% to try to fit this transmission spectrum. 00:16:18.040 --> 00:16:19.660 align:middle line:90% This is what I end up getting. 00:16:19.660 --> 00:16:22.820 align:middle line:84% So the vertical axis is essentially wavelength. 00:16:22.820 --> 00:16:25.160 align:middle line:84% So what this does, what you are seeing here, 00:16:25.160 --> 00:16:29.320 align:middle line:84% they correspond to the extremum positions and the transmission 00:16:29.320 --> 00:16:30.960 align:middle line:90% spectrum, the maximum/minimum. 00:16:30.960 --> 00:16:32.513 align:middle line:90% I extracted it from that plot. 00:16:32.513 --> 00:16:34.180 align:middle line:84% And essentially, I just did the fitting. 00:16:34.180 --> 00:16:36.200 align:middle line:84% So if everything falls on one line, 00:16:36.200 --> 00:16:40.220 align:middle line:84% it says that this is indeed caused by reflectance. 00:16:40.220 --> 00:16:42.520 align:middle line:84% I basically fitted to this very simple model there. 00:16:42.520 --> 00:16:47.720 align:middle line:84% OK, so while we still don't know exactly why you 00:16:47.720 --> 00:16:49.820 align:middle line:84% have this scratch-resistant property, 00:16:49.820 --> 00:16:52.393 align:middle line:84% we can at least conclude, based on my fitting, 00:16:52.393 --> 00:16:54.560 align:middle line:84% that there's actually a surface coating layer, which 00:16:54.560 --> 00:16:58.040 align:middle line:84% is about 2.3 microns in thickness 00:16:58.040 --> 00:17:00.880 align:middle line:84% and has a refraction index 0.03 higher or lower 00:17:00.880 --> 00:17:02.360 align:middle line:90% compared to a substrate glass. 00:17:02.360 --> 00:17:05.580 align:middle line:84% So it's plausible that this coating is also 00:17:05.580 --> 00:17:08.040 align:middle line:84% a coating with a higher hardness, 00:17:08.040 --> 00:17:10.740 align:middle line:84% so they prevent essentially any scratches on the glass surface. 00:17:10.740 --> 00:17:12.942 align:middle line:84% Again, this is speculation, but it makes sense 00:17:12.942 --> 00:17:14.359 align:middle line:84% because that's the only difference 00:17:14.359 --> 00:17:17.200 align:middle line:84% I see between this glass material and the untreated 00:17:17.200 --> 00:17:17.819 align:middle line:90% Gorilla. 00:17:17.819 --> 00:17:21.435 align:middle line:90% 00:17:21.435 --> 00:17:23.060 align:middle line:84% Now let's take a look at another trick. 00:17:23.060 --> 00:17:26.663 align:middle line:84% This one I think is much more ingenious. 00:17:26.663 --> 00:17:29.080 align:middle line:84% This actually does engineer the glass intrinsic properties 00:17:29.080 --> 00:17:30.980 align:middle line:90% to get the damage resistance. 00:17:30.980 --> 00:17:33.300 align:middle line:84% So this is another line of products. 00:17:33.300 --> 00:17:36.080 align:middle line:84% They call it native glass materials with native damage 00:17:36.080 --> 00:17:37.400 align:middle line:90% resistance, NDR. 00:17:37.400 --> 00:17:40.380 align:middle line:84% So at the bottom, again I'm showing you what happens here. 00:17:40.380 --> 00:17:42.160 align:middle line:84% If you just scratch on glass material 00:17:42.160 --> 00:17:44.880 align:middle line:84% without this capability, you get large scratches. 00:17:44.880 --> 00:17:48.360 align:middle line:84% With this capability, the scratch is more confined. 00:17:48.360 --> 00:17:50.300 align:middle line:84% Microscopically, if you look at what happens, 00:17:50.300 --> 00:17:53.480 align:middle line:84% this is a cross section of the scratch. 00:17:53.480 --> 00:17:56.900 align:middle line:84% On other materials without this capability, 00:17:56.900 --> 00:17:58.758 align:middle line:84% you see lots of microcracks forming. 00:17:58.758 --> 00:18:01.300 align:middle line:84% And that happens because when you actually try to scratch it, 00:18:01.300 --> 00:18:04.080 align:middle line:84% you apply this essentially compressive stress. 00:18:04.080 --> 00:18:06.390 align:middle line:84% Then, stress accumulation exceeds the strength 00:18:06.390 --> 00:18:08.640 align:middle line:84% of the material at the surface, which eventually leads 00:18:08.640 --> 00:18:10.760 align:middle line:90% to a formation of microcracks. 00:18:10.760 --> 00:18:13.760 align:middle line:84% On the other hand, if it is glass with NDR, 00:18:13.760 --> 00:18:17.000 align:middle line:84% you see this region here, this part, this 00:18:17.000 --> 00:18:21.200 align:middle line:84% is the region with a lot of plastic deformation. 00:18:21.200 --> 00:18:26.160 align:middle line:84% Essentially, the cracks is inhibited 00:18:26.160 --> 00:18:28.520 align:middle line:84% because the plastic formation helps to absorb 00:18:28.520 --> 00:18:31.480 align:middle line:90% a lot of the deformation. 00:18:31.480 --> 00:18:34.340 align:middle line:84% So this is a glass that can induce a considerable amount 00:18:34.340 --> 00:18:35.840 align:middle line:84% of plastic deformation, essentially, 00:18:35.840 --> 00:18:37.600 align:middle line:90% in the form of densification. 00:18:37.600 --> 00:18:40.360 align:middle line:90% 00:18:40.360 --> 00:18:41.620 align:middle line:90% So how does it work? 00:18:41.620 --> 00:18:48.000 align:middle line:90% 00:18:48.000 --> 00:18:50.040 align:middle line:84% So the trick goes back to when we talk 00:18:50.040 --> 00:18:52.680 align:middle line:90% about borosilicate glasses. 00:18:52.680 --> 00:18:54.520 align:middle line:84% So when we talk about borosilicate glasses, 00:18:54.520 --> 00:18:57.568 align:middle line:84% we mentioned that boron has this boron anomaly. 00:18:57.568 --> 00:18:59.360 align:middle line:84% It can essentially switch between two kinds 00:18:59.360 --> 00:19:01.080 align:middle line:90% of microstructure moieties. 00:19:01.080 --> 00:19:04.800 align:middle line:84% It can either be forming essentially a B3. 00:19:04.800 --> 00:19:06.040 align:middle line:90% I show them in there. 00:19:06.040 --> 00:19:08.740 align:middle line:84% You can actually have a non-bridging oxygen. 00:19:08.740 --> 00:19:11.660 align:middle line:84% Or it can induce this B3 to B4 conversion. 00:19:11.660 --> 00:19:14.400 align:middle line:84% You form a fourfold coordinate boron atom which give rise 00:19:14.400 --> 00:19:18.480 align:middle line:84% to a more tightly connected network, which gives you 00:19:18.480 --> 00:19:22.240 align:middle line:84% essentially a higher density, among other properties, 00:19:22.240 --> 00:19:25.280 align:middle line:84% like lower thermal expansion coefficients. 00:19:25.280 --> 00:19:29.560 align:middle line:84% So the trick they play is that they essentially add boron 00:19:29.560 --> 00:19:31.350 align:middle line:90% into the glass network. 00:19:31.350 --> 00:19:33.790 align:middle line:84% And they engineer chemistry so that when you actually 00:19:33.790 --> 00:19:36.430 align:middle line:84% apply stress, the stress actually 00:19:36.430 --> 00:19:41.190 align:middle line:84% induces a transformation from B3 to B4. 00:19:41.190 --> 00:19:44.070 align:middle line:84% In the process, the density of glass 00:19:44.070 --> 00:19:47.710 align:middle line:84% increases because of increased network connectivity. 00:19:47.710 --> 00:19:50.050 align:middle line:84% And this is the plastic deformation. 00:19:50.050 --> 00:19:53.670 align:middle line:84% It's a permanent plastic deformation. 00:19:53.670 --> 00:19:56.030 align:middle line:84% So it's a stress-induced chemical transformation which 00:19:56.030 --> 00:19:57.930 align:middle line:90% also brings in a densification. 00:19:57.930 --> 00:20:00.830 align:middle line:90% 00:20:00.830 --> 00:20:04.350 align:middle line:84% So that's why this material can now actually absorb 00:20:04.350 --> 00:20:06.830 align:middle line:84% a lot more deformation while inhibiting 00:20:06.830 --> 00:20:10.952 align:middle line:84% the formation of cracks, which I think is a pretty smart trick. 00:20:10.952 --> 00:20:12.910 align:middle line:84% They actually use the structural transformation 00:20:12.910 --> 00:20:15.410 align:middle line:84% to help prevent the crack formation. 00:20:15.410 --> 00:20:18.894 align:middle line:90% 00:20:18.894 --> 00:20:22.630 align:middle line:84% Any questions about that before I move on? 00:20:22.630 --> 00:20:23.210 align:middle line:90% OK. 00:20:23.210 --> 00:20:30.210 align:middle line:84% So now, besides being scratch resistant or damage resistant, 00:20:30.210 --> 00:20:33.270 align:middle line:84% I want to mention that the other advances that people 00:20:33.270 --> 00:20:37.950 align:middle line:84% were able to accomplish is to make bendable phone screens. 00:20:37.950 --> 00:20:42.190 align:middle line:84% So this is actually from a news release by Samsung. 00:20:42.190 --> 00:20:45.190 align:middle line:84% I think this is about 2020 or something. 00:20:45.190 --> 00:20:46.910 align:middle line:84% They essentially unveil what they 00:20:46.910 --> 00:20:50.470 align:middle line:84% claim is the first foldable phone with a glass screen, the Z 00:20:50.470 --> 00:20:52.550 align:middle line:90% Flip. 00:20:52.550 --> 00:20:55.230 align:middle line:84% In the news release, you can see that you can essentially 00:20:55.230 --> 00:20:56.590 align:middle line:90% fold the phone in half. 00:20:56.590 --> 00:21:01.870 align:middle line:84% They claim that they were able to bend the laws of physics 00:21:01.870 --> 00:21:05.030 align:middle line:90% to get glasses to be bendable. 00:21:05.030 --> 00:21:06.368 align:middle line:90% Is that really the case? 00:21:06.368 --> 00:21:07.410 align:middle line:90% Well, we can take a look. 00:21:07.410 --> 00:21:11.452 align:middle line:84% First, actually, I'm going to show you a little YouTube video. 00:21:11.452 --> 00:21:12.910 align:middle line:84% The key trick that you can actually 00:21:12.910 --> 00:21:15.870 align:middle line:84% get these glasses to be really flexible and foldable 00:21:15.870 --> 00:21:17.590 align:middle line:90% is to make them extremely thin. 00:21:17.590 --> 00:21:23.510 align:middle line:90% 00:21:23.510 --> 00:21:24.570 align:middle line:90% This one has no sound. 00:21:24.570 --> 00:21:27.710 align:middle line:90% [VIDEO PLAYBACK] 00:21:27.710 --> 00:21:30.990 align:middle line:90% 00:21:30.990 --> 00:21:32.570 align:middle line:84% Internet connection is not very good. 00:21:32.570 --> 00:21:33.070 align:middle line:90% Oops. 00:21:33.070 --> 00:21:41.650 align:middle line:90% 00:21:41.650 --> 00:21:45.070 align:middle line:84% So this company specializes in laser processing 00:21:45.070 --> 00:21:46.728 align:middle line:90% of these ultra-thin glasses. 00:21:46.728 --> 00:21:48.270 align:middle line:84% So you can see that they can actually 00:21:48.270 --> 00:21:50.870 align:middle line:84% use the surface patterns to help relieve the stress. 00:21:50.870 --> 00:21:53.412 align:middle line:84% They have actually some data to show that. 00:21:53.412 --> 00:21:54.870 align:middle line:84% You can see the stress distribution 00:21:54.870 --> 00:21:56.430 align:middle line:84% with the surface patterning helping to release 00:21:56.430 --> 00:21:57.610 align:middle line:90% the stress at the edges. 00:21:57.610 --> 00:22:02.050 align:middle line:84% But they were able to bend it to a very small bending radius. 00:22:02.050 --> 00:22:03.570 align:middle line:90% And it's completely flexible. 00:22:03.570 --> 00:22:05.070 align:middle line:90% [END PLAYBACK] 00:22:05.070 --> 00:22:06.910 align:middle line:90% So what's the physics here? 00:22:06.910 --> 00:22:10.070 align:middle line:84% Are we actually breaking any physical laws? 00:22:10.070 --> 00:22:13.990 align:middle line:84% If you think about if you bend a sheet of glass, 00:22:13.990 --> 00:22:15.870 align:middle line:90% that's mechanics 101. 00:22:15.870 --> 00:22:18.030 align:middle line:90% What happens? 00:22:18.030 --> 00:22:21.870 align:middle line:84% On the top surface, you have essentially tensile strain. 00:22:21.870 --> 00:22:25.350 align:middle line:84% At the bottom surface, you get compressive strain. 00:22:25.350 --> 00:22:29.230 align:middle line:84% If the glass is uniform, then right in the middle, 00:22:29.230 --> 00:22:33.030 align:middle line:84% you have what's called a neutral plane or neutral axis, 00:22:33.030 --> 00:22:38.470 align:middle line:84% where the stress essentially goes to 0. 00:22:38.470 --> 00:22:41.230 align:middle line:84% So if a glass were to break, this top surface 00:22:41.230 --> 00:22:45.190 align:middle line:84% where you have tensile stress is usually where any kind of defect 00:22:45.190 --> 00:22:46.630 align:middle line:90% is going to initiate. 00:22:46.630 --> 00:22:53.390 align:middle line:84% So the strain level on the top surface, the strain, epsilon, 00:22:53.390 --> 00:22:57.370 align:middle line:84% on the top surface, this is actually given by-- basically, 00:22:57.370 --> 00:23:01.870 align:middle line:84% if we label the glass thickness, this is, let's say, d; 00:23:01.870 --> 00:23:06.950 align:middle line:84% the radius of the bending here, let's label it as r-- 00:23:06.950 --> 00:23:13.670 align:middle line:84% then, the strain on the top surface will just be given 00:23:13.670 --> 00:23:20.150 align:middle line:84% by epsilon equals to 1/2 thickness normalized 00:23:20.150 --> 00:23:21.813 align:middle line:90% by the bending radius. 00:23:21.813 --> 00:23:23.230 align:middle line:84% That's just a very simple equation 00:23:23.230 --> 00:23:24.750 align:middle line:84% for mechanics that gives you what's 00:23:24.750 --> 00:23:26.560 align:middle line:90% the strain on the top surface. 00:23:26.560 --> 00:23:31.052 align:middle line:90% 00:23:31.052 --> 00:23:32.820 align:middle line:84% So now let's plug in some numbers 00:23:32.820 --> 00:23:34.160 align:middle line:90% and see what the strain is. 00:23:34.160 --> 00:23:37.460 align:middle line:84% And from there, we can calculate the stress. 00:23:37.460 --> 00:23:40.180 align:middle line:84% So for these ultra-thin glasses, their thickness 00:23:40.180 --> 00:23:43.700 align:middle line:84% is of the order of 25 to 30 microns. 00:23:43.700 --> 00:23:46.120 align:middle line:84% Let's just take, let's say, 30 microns. 00:23:46.120 --> 00:23:52.340 align:middle line:84% So 30 microns is 0.03 millimeters. 00:23:52.340 --> 00:23:53.820 align:middle line:84% In terms of bending radius, we see 00:23:53.820 --> 00:23:56.540 align:middle line:84% that they can be bent to a very tight bending radius, down 00:23:56.540 --> 00:23:57.980 align:middle line:90% to about 1 millimeter. 00:23:57.980 --> 00:23:59.940 align:middle line:84% So let's just take a rounded number. 00:23:59.940 --> 00:24:04.580 align:middle line:84% This is 2 times 1 millimeter gives you 2 millimeters. 00:24:04.580 --> 00:24:07.860 align:middle line:90% 00:24:07.860 --> 00:24:10.740 align:middle line:84% So it's pretty straightforward to calculate that the strain now 00:24:10.740 --> 00:24:15.140 align:middle line:90% on the surface is 1.5%. 00:24:15.140 --> 00:24:16.940 align:middle line:90% Now what about the stress? 00:24:16.940 --> 00:24:19.860 align:middle line:84% So we need to know the Young's modulus if we want 00:24:19.860 --> 00:24:21.760 align:middle line:90% to infer what the stress is. 00:24:21.760 --> 00:24:23.020 align:middle line:90% So come back here. 00:24:23.020 --> 00:24:26.720 align:middle line:84% The Young's modulus for glass, 60 to 75 gigapascals. 00:24:26.720 --> 00:24:29.460 align:middle line:90% Let's just say 60 gigapascals. 00:24:29.460 --> 00:24:37.080 align:middle line:84% Then the stress is going to be E times epsilon. 00:24:37.080 --> 00:24:41.460 align:middle line:90% 00:24:41.460 --> 00:24:45.980 align:middle line:90% So what does that correspond to? 00:24:45.980 --> 00:24:48.120 align:middle line:90% So it will be 60 times 1.5. 00:24:48.120 --> 00:24:59.520 align:middle line:84% So it becomes a 0.9 gigapascals, or 900 megapascals. 00:24:59.520 --> 00:25:04.500 align:middle line:90% 00:25:04.500 --> 00:25:06.500 align:middle line:84% So if you have a glass sheet that has not 00:25:06.500 --> 00:25:08.020 align:middle line:84% been chemically strengthened, this 00:25:08.020 --> 00:25:11.920 align:middle line:84% would be the level of tensile stress the glass will see. 00:25:11.920 --> 00:25:16.420 align:middle line:84% However, you notice that this glass material, 00:25:16.420 --> 00:25:17.920 align:middle line:84% once they are properly strengthened, 00:25:17.920 --> 00:25:21.220 align:middle line:84% they can have a compressive stress of the order of 900 00:25:21.220 --> 00:25:22.940 align:middle line:90% megapascals or above. 00:25:22.940 --> 00:25:26.420 align:middle line:84% So it means even under these pretty extreme bending 00:25:26.420 --> 00:25:29.820 align:middle line:84% conditions, the surface is still almost 00:25:29.820 --> 00:25:32.700 align:middle line:84% in a tensile stress-free state or even 00:25:32.700 --> 00:25:34.360 align:middle line:84% still in a compressive stress state. 00:25:34.360 --> 00:25:40.500 align:middle line:90% 00:25:40.500 --> 00:25:42.480 align:middle line:84% Because of the native compressive stress, 00:25:42.480 --> 00:25:45.780 align:middle line:84% you have to add it on top of this stress here. 00:25:45.780 --> 00:25:48.580 align:middle line:84% So if you're under neutral condition 00:25:48.580 --> 00:25:50.240 align:middle line:84% or under compressive stress condition, 00:25:50.240 --> 00:25:53.460 align:middle line:84% obviously you're not going to get a lot of crack propagation. 00:25:53.460 --> 00:25:56.820 align:middle line:84% So that's why we see that there's really 00:25:56.820 --> 00:26:00.380 align:middle line:90% no breaking of physical laws. 00:26:00.380 --> 00:26:02.100 align:middle line:84% The whole idea is that as long as you 00:26:02.100 --> 00:26:04.140 align:middle line:84% make the glass material thin enough 00:26:04.140 --> 00:26:06.340 align:middle line:84% and also apply large enough surface 00:26:06.340 --> 00:26:08.740 align:middle line:84% compressive stress through chemical strengthening, 00:26:08.740 --> 00:26:11.700 align:middle line:84% you will be able to prevent any crack initiation, even 00:26:11.700 --> 00:26:13.640 align:middle line:84% under these predicted bending conditions. 00:26:13.640 --> 00:26:20.970 align:middle line:90% 00:26:20.970 --> 00:26:24.580 align:middle line:84% So the whole idea is to get a very thin piece of glass that 00:26:24.580 --> 00:26:27.140 align:middle line:84% has a thickness that's 30 microns, 00:26:27.140 --> 00:26:29.360 align:middle line:90% about 1/3 of a sheet of paper. 00:26:29.360 --> 00:26:30.800 align:middle line:90% So it's really very, very thin. 00:26:30.800 --> 00:26:38.687 align:middle line:90% 00:26:38.687 --> 00:26:40.520 align:middle line:84% Now, one thing I do want to mention, though, 00:26:40.520 --> 00:26:42.160 align:middle line:84% about this particular phone here, 00:26:42.160 --> 00:26:46.340 align:middle line:84% Z Flip, if you look at the way the process is demonstrated 00:26:46.340 --> 00:26:48.880 align:middle line:84% here, the screen is actually not one single-layer material. 00:26:48.880 --> 00:26:50.338 align:middle line:84% They actually put two layers there, 00:26:50.338 --> 00:26:52.020 align:middle line:84% essentially a glass that they also 00:26:52.020 --> 00:26:54.900 align:middle line:84% put on top a polymer layer as a protective coating 00:26:54.900 --> 00:26:57.140 align:middle line:90% to prevent surface damage. 00:26:57.140 --> 00:26:59.620 align:middle line:84% Downside is, polymer, of course, is soft. 00:26:59.620 --> 00:27:02.485 align:middle line:84% It's not scratch resistant, even compared to glass. 00:27:02.485 --> 00:27:03.860 align:middle line:84% And that's why the phone actually 00:27:03.860 --> 00:27:06.660 align:middle line:84% gets quite a bit of criticism because of the poor ability 00:27:06.660 --> 00:27:08.400 align:middle line:84% to withstand the surface scratches. 00:27:08.400 --> 00:27:10.200 align:middle line:90% But that's the trade off there. 00:27:10.200 --> 00:27:13.180 align:middle line:90% 00:27:13.180 --> 00:27:20.130 align:middle line:84% So now, let's move beyond cover materials and let's look at, 00:27:20.130 --> 00:27:24.690 align:middle line:84% these are glass materials with extremely high K steps. 00:27:24.690 --> 00:27:26.450 align:middle line:84% You may recall that K steps is basically 00:27:26.450 --> 00:27:28.970 align:middle line:84% the diffusion distance of the ions during the ion exchange 00:27:28.970 --> 00:27:29.690 align:middle line:90% process. 00:27:29.690 --> 00:27:33.690 align:middle line:84% So these are glass materials with very high fracture 00:27:33.690 --> 00:27:36.970 align:middle line:84% resistance compared to typical materials because of large K 00:27:36.970 --> 00:27:39.850 align:middle line:84% steps and extremely high level of surface stress. 00:27:39.850 --> 00:27:42.290 align:middle line:84% So these are the materials that are 00:27:42.290 --> 00:27:46.370 align:middle line:84% used for some high-end applications, 00:27:46.370 --> 00:27:48.330 align:middle line:84% for architectural applications, for example. 00:27:48.330 --> 00:27:49.930 align:middle line:84% They are also the materials people use 00:27:49.930 --> 00:27:52.802 align:middle line:84% on essentially military vehicles as the window materials. 00:27:52.802 --> 00:27:55.510 align:middle line:84% They can actually be bulletproof because of very, very high level 00:27:55.510 --> 00:27:56.550 align:middle line:90% of surface stress. 00:27:56.550 --> 00:27:58.690 align:middle line:84% So let's take a look at the chemistry 00:27:58.690 --> 00:28:01.090 align:middle line:84% here and see how they were able to achieve 00:28:01.090 --> 00:28:02.950 align:middle line:90% these extremely high K steps. 00:28:02.950 --> 00:28:04.890 align:middle line:90% Look at the K steps here. 00:28:04.890 --> 00:28:08.130 align:middle line:84% These are typically hundreds of microns K steps. 00:28:08.130 --> 00:28:12.650 align:middle line:84% And in the extreme case, you can even get 1-millimeter-thick K 00:28:12.650 --> 00:28:15.370 align:middle line:90% steps. 00:28:15.370 --> 00:28:19.050 align:middle line:84% So to get to this high K steps, these are all 00:28:19.050 --> 00:28:22.210 align:middle line:90% aluminum silicate glasses. 00:28:22.210 --> 00:28:24.210 align:middle line:84% As we mentioned, you have alumina there 00:28:24.210 --> 00:28:27.170 align:middle line:84% because you want to essentially boost the glass transition 00:28:27.170 --> 00:28:30.150 align:middle line:84% temperature so you don't get glass relaxation even if you do 00:28:30.150 --> 00:28:31.490 align:middle line:90% a high temperature treatment. 00:28:31.490 --> 00:28:34.250 align:middle line:84% You can use high temperature to expedite the ionic diffusion 00:28:34.250 --> 00:28:34.850 align:middle line:90% process. 00:28:34.850 --> 00:28:37.690 align:middle line:84% So that's why you see that pretty consistently, they 00:28:37.690 --> 00:28:38.850 align:middle line:90% all contain alumina. 00:28:38.850 --> 00:28:43.010 align:middle line:84% Also, the other thing is that you notice that they always 00:28:43.010 --> 00:28:46.210 align:middle line:84% contain a couple of percent or few percent of magnesium oxide 00:28:46.210 --> 00:28:48.080 align:middle line:90% or calcium oxide. 00:28:48.080 --> 00:28:49.830 align:middle line:84% This is what we mentioned in last lecture. 00:28:49.830 --> 00:28:51.247 align:middle line:84% Empirically, people have found out 00:28:51.247 --> 00:28:52.810 align:middle line:84% that adding this rare-earth oxide 00:28:52.810 --> 00:28:55.110 align:middle line:84% helps the diffusion of alkaline ions. 00:28:55.110 --> 00:28:57.710 align:middle line:90% 00:28:57.710 --> 00:29:02.450 align:middle line:84% And then in terms of the ion exchange process, 00:29:02.450 --> 00:29:04.650 align:middle line:84% you also notice that they use a mixed ion 00:29:04.650 --> 00:29:07.303 align:middle line:84% bath for these ones with very high K steps. 00:29:07.303 --> 00:29:09.470 align:middle line:84% This is also something we mentioned in last lecture. 00:29:09.470 --> 00:29:11.810 align:middle line:84% If you do a mixed ion bath, this will actually 00:29:11.810 --> 00:29:14.370 align:middle line:84% also help to increase the diffusion of the ions 00:29:14.370 --> 00:29:15.390 align:middle line:90% into the surface. 00:29:15.390 --> 00:29:19.890 align:middle line:84% Now, despite all this treatment, this process still 00:29:19.890 --> 00:29:23.290 align:middle line:84% takes a very long time just because of the kinetics 00:29:23.290 --> 00:29:25.330 align:middle line:90% of diffusion process. 00:29:25.330 --> 00:29:30.210 align:middle line:84% So look at the time it takes to process the glass materials. 00:29:30.210 --> 00:29:36.530 align:middle line:84% They range from one day to the case of half a month for the ion 00:29:36.530 --> 00:29:37.497 align:middle line:90% strengthening process. 00:29:37.497 --> 00:29:39.330 align:middle line:84% They just allow ion to essentially penetrate 00:29:39.330 --> 00:29:42.610 align:middle line:84% through this large surface thickness. 00:29:42.610 --> 00:29:45.130 align:middle line:84% So that's an extreme example where 00:29:45.130 --> 00:29:47.770 align:middle line:84% you can actually get to this very high strength glass 00:29:47.770 --> 00:29:48.487 align:middle line:90% materials. 00:29:48.487 --> 00:29:50.570 align:middle line:84% Now these are actually some quantitative data just 00:29:50.570 --> 00:29:52.912 align:middle line:84% to show you that in this first plot here, 00:29:52.912 --> 00:29:55.370 align:middle line:84% this is just to illustrate that this kind of mixed ion bath 00:29:55.370 --> 00:29:58.810 align:middle line:84% actually helps with promoting the diffusion kinetics. 00:29:58.810 --> 00:30:00.230 align:middle line:90% So you see these dots here. 00:30:00.230 --> 00:30:03.240 align:middle line:90% 00:30:03.240 --> 00:30:06.270 align:middle line:84% You can see from the horizontal axis here, 00:30:06.270 --> 00:30:08.050 align:middle line:84% this is basically the depth below surface 00:30:08.050 --> 00:30:09.390 align:middle line:90% where the ion has diffused into. 00:30:09.390 --> 00:30:12.170 align:middle line:84% You can clearly see that if you mix into a mixed ion 00:30:12.170 --> 00:30:15.570 align:middle line:84% bath with sodium or potassium, you get more penetration 00:30:15.570 --> 00:30:17.770 align:middle line:84% into the glass material compared to the ones using 00:30:17.770 --> 00:30:19.290 align:middle line:90% only one kind of ionic species. 00:30:19.290 --> 00:30:22.275 align:middle line:90% 00:30:22.275 --> 00:30:24.130 align:middle line:90% Weibull plot again. 00:30:24.130 --> 00:30:26.690 align:middle line:84% So on the Weibull plot, we also see that we 00:30:26.690 --> 00:30:29.490 align:middle line:90% can get much higher strength. 00:30:29.490 --> 00:30:31.038 align:middle line:84% These are the points that actually 00:30:31.038 --> 00:30:32.830 align:middle line:84% correspond to strengthened glass materials. 00:30:32.830 --> 00:30:35.330 align:middle line:84% These are the ones that are actually not strengthened. 00:30:35.330 --> 00:30:39.370 align:middle line:84% With mixed ion bath, these are the ones with the highest 00:30:39.370 --> 00:30:40.750 align:middle line:90% strength, the green dots. 00:30:40.750 --> 00:30:42.550 align:middle line:84% These are actually with the mixed ion bath. 00:30:42.550 --> 00:30:44.090 align:middle line:90% So they're high strength. 00:30:44.090 --> 00:30:48.970 align:middle line:84% At the same time, you also see the same thing, improved Weibull 00:30:48.970 --> 00:30:51.330 align:middle line:84% modulus, indicating that the material after ion exchange 00:30:51.330 --> 00:30:53.550 align:middle line:84% becomes more consistent, in terms of quality. 00:30:53.550 --> 00:30:57.690 align:middle line:90% 00:30:57.690 --> 00:31:01.673 align:middle line:84% All right, so these are for essentially flat-panel glasses, 00:31:01.673 --> 00:31:03.090 align:middle line:84% how you use chemical strengthening 00:31:03.090 --> 00:31:06.050 align:middle line:84% to increase its ability to resist fracturing. 00:31:06.050 --> 00:31:09.210 align:middle line:84% The other case where ion or chemical strengthening or ion 00:31:09.210 --> 00:31:13.520 align:middle line:84% exchange is widely used is for glass containers, 00:31:13.520 --> 00:31:16.600 align:middle line:84% especially vials used for pharmaceutical applications. 00:31:16.600 --> 00:31:19.440 align:middle line:84% This is one case where you cannot do conventional thermal 00:31:19.440 --> 00:31:20.520 align:middle line:90% tempering. 00:31:20.520 --> 00:31:24.020 align:middle line:84% Because of the geometry, inherent strange geometry 00:31:24.020 --> 00:31:25.840 align:middle line:84% and non-uniform thickness of a vial, 00:31:25.840 --> 00:31:28.120 align:middle line:84% if you do thermal tempering, it's 00:31:28.120 --> 00:31:30.480 align:middle line:84% going to lead to a very non-uniform cooling profile, 00:31:30.480 --> 00:31:32.620 align:middle line:84% therefore leading to non-uniform thermal stress. 00:31:32.620 --> 00:31:36.840 align:middle line:84% So that's why people basically use a chemical toughening 00:31:36.840 --> 00:31:41.200 align:middle line:84% to strengthen these vials for pharmaceutical applications. 00:31:41.200 --> 00:31:43.000 align:middle line:84% So what I'm quoting here, this is actually 00:31:43.000 --> 00:31:46.920 align:middle line:84% an article that's from 2020, essentially 00:31:46.920 --> 00:31:50.600 align:middle line:84% at the peak of COVID, where people are actually 00:31:50.600 --> 00:31:54.760 align:middle line:84% racing to develop vaccines against COVID. 00:31:54.760 --> 00:31:58.240 align:middle line:84% One of the issues they faced there is not just 00:31:58.240 --> 00:32:00.920 align:middle line:84% to scale up the manufacturing of the vaccine itself, 00:32:00.920 --> 00:32:04.760 align:middle line:84% but also the container that actually contain those vaccines. 00:32:04.760 --> 00:32:07.120 align:middle line:84% And this is not a small problem because you 00:32:07.120 --> 00:32:11.360 align:middle line:84% need billions of glass containers 00:32:11.360 --> 00:32:15.880 align:middle line:84% so that they were able to ship this COVID virus 00:32:15.880 --> 00:32:18.040 align:middle line:90% products to around the world. 00:32:18.040 --> 00:32:20.180 align:middle line:84% The requirement is also pretty stringent here. 00:32:20.180 --> 00:32:24.130 align:middle line:90% 00:32:24.130 --> 00:32:27.660 align:middle line:84% For this vials used to contain vaccines, 00:32:27.660 --> 00:32:30.440 align:middle line:84% they need to withstand a very wide temperature range. 00:32:30.440 --> 00:32:33.600 align:middle line:84% You may recall that a lot of the vaccines, they 00:32:33.600 --> 00:32:35.760 align:middle line:84% had to be stored at cryogenic temperatures. 00:32:35.760 --> 00:32:37.760 align:middle line:84% For example, Pfizer's vaccine, they 00:32:37.760 --> 00:32:39.960 align:middle line:84% have to store at negative 70 degrees Celsius. 00:32:39.960 --> 00:32:42.720 align:middle line:84% They also need to withstand high temperatures 00:32:42.720 --> 00:32:46.200 align:middle line:84% because these containers need to be sterilized. 00:32:46.200 --> 00:32:48.690 align:middle line:84% And they usually sterilized at high temperature, 200, 300 00:32:48.690 --> 00:32:49.840 align:middle line:90% degrees Celsius. 00:32:49.840 --> 00:32:53.100 align:middle line:84% So unlike everyday-use glass material, 00:32:53.100 --> 00:32:56.520 align:middle line:84% which usually sees a much narrower temperature window, all 00:32:56.520 --> 00:32:58.040 align:middle line:84% the pharmaceutical vials, they have 00:32:58.040 --> 00:33:02.360 align:middle line:84% to be specified to operate from negative 100 degrees Celsius 00:33:02.360 --> 00:33:06.280 align:middle line:84% to positive 400 degrees Celsius, which is a very wide temperature 00:33:06.280 --> 00:33:08.300 align:middle line:84% range compared to other kind of glass products. 00:33:08.300 --> 00:33:11.960 align:middle line:90% 00:33:11.960 --> 00:33:15.920 align:middle line:84% They also operate in a potentially chemically corrosive 00:33:15.920 --> 00:33:16.540 align:middle line:90% environment. 00:33:16.540 --> 00:33:18.040 align:middle line:84% They have been in contact with aqueous solution 00:33:18.040 --> 00:33:19.293 align:middle line:90% for a long period of time. 00:33:19.293 --> 00:33:20.960 align:middle line:84% So you don't want any kind of components 00:33:20.960 --> 00:33:23.400 align:middle line:84% from the glass to actually leach into the solution. 00:33:23.400 --> 00:33:28.560 align:middle line:84% They also have to have a very high resistance 00:33:28.560 --> 00:33:34.440 align:middle line:84% to surface damage, essentially resistant to wear and tear. 00:33:34.440 --> 00:33:36.560 align:middle line:84% And most of their requirements actually 00:33:36.560 --> 00:33:38.760 align:middle line:84% come from the manufacturing side. 00:33:38.760 --> 00:33:41.280 align:middle line:84% You can imagine that if you have a very delicate glass 00:33:41.280 --> 00:33:43.160 align:middle line:84% product, then you have to be very, very 00:33:43.160 --> 00:33:46.400 align:middle line:84% careful in handling them during manufacturing 00:33:46.400 --> 00:33:47.680 align:middle line:90% and during packaging. 00:33:47.680 --> 00:33:51.760 align:middle line:84% So that actually lowers the throughput. 00:33:51.760 --> 00:33:54.200 align:middle line:84% So to use high throughput process, 00:33:54.200 --> 00:33:57.220 align:middle line:84% they also require the glass with very high surface strength. 00:33:57.220 --> 00:34:02.594 align:middle line:90% 00:34:02.594 --> 00:34:06.105 align:middle line:84% So that's why-- this is one example-- 00:34:06.105 --> 00:34:08.480 align:middle line:84% for example, Corning actually developed this special line 00:34:08.480 --> 00:34:10.264 align:middle line:90% of glass they call Valor Glass. 00:34:10.264 --> 00:34:11.639 align:middle line:84% It has several important features 00:34:11.639 --> 00:34:13.239 align:middle line:84% that make it suited for this particular kind 00:34:13.239 --> 00:34:14.800 align:middle line:84% of pharmaceutical applications here. 00:34:14.800 --> 00:34:17.420 align:middle line:84% The first one is that it's boron-free. 00:34:17.420 --> 00:34:20.360 align:middle line:90% 00:34:20.360 --> 00:34:23.040 align:middle line:84% Now we have mentioned about boron, essentially 00:34:23.040 --> 00:34:24.159 align:middle line:90% like Pyrex glass. 00:34:24.159 --> 00:34:26.903 align:middle line:84% Pyrex glass is used in a lot of these applications 00:34:26.903 --> 00:34:28.320 align:middle line:84% when you're dealing with chemicals 00:34:28.320 --> 00:34:31.480 align:middle line:84% because of essentially you have this tightly connected network. 00:34:31.480 --> 00:34:33.260 align:middle line:84% It has good thermal shock resistance. 00:34:33.260 --> 00:34:35.960 align:middle line:84% It's also more chemically resistant. 00:34:35.960 --> 00:34:40.280 align:middle line:84% However, when it comes to pharmaceutical vials, 00:34:40.280 --> 00:34:42.639 align:middle line:84% people don't like to use boron because it 00:34:42.639 --> 00:34:47.040 align:middle line:84% can cause a problem of so-called delamination. 00:34:47.040 --> 00:34:48.400 align:middle line:90% So why do you get delamination? 00:34:48.400 --> 00:34:50.400 align:middle line:84% First thing, it has to do with the manufacturing 00:34:50.400 --> 00:34:51.719 align:middle line:90% process of these vials. 00:34:51.719 --> 00:34:54.120 align:middle line:84% So if you have a regular glass bottle, the way 00:34:54.120 --> 00:34:55.848 align:middle line:90% you make it is by blow molding. 00:34:55.848 --> 00:34:57.640 align:middle line:84% You blow compressed air through the parison 00:34:57.640 --> 00:34:59.920 align:middle line:84% and it conforms to the shape to get the bottle. 00:34:59.920 --> 00:35:03.640 align:middle line:84% Blow molding is not used for this pharmaceutical vials 00:35:03.640 --> 00:35:04.420 align:middle line:90% for a few reasons. 00:35:04.420 --> 00:35:07.372 align:middle line:84% First thing, blow molding, borosilicate glass typically 00:35:07.372 --> 00:35:09.330 align:middle line:84% has a much higher glass transition temperature, 00:35:09.330 --> 00:35:11.170 align:middle line:84% so it means the processing temperature is a lot higher. 00:35:11.170 --> 00:35:13.010 align:middle line:84% So it makes it a more challenging process. 00:35:13.010 --> 00:35:16.150 align:middle line:84% Second is that blow molding typically 00:35:16.150 --> 00:35:17.910 align:middle line:84% causes very large plastic deformation 00:35:17.910 --> 00:35:19.868 align:middle line:84% or very large deformation, viscous deformation, 00:35:19.868 --> 00:35:21.150 align:middle line:90% during the process. 00:35:21.150 --> 00:35:23.390 align:middle line:84% Large deformation means that there's also 00:35:23.390 --> 00:35:26.690 align:middle line:84% a much higher residual stress, as you can see from the bottle. 00:35:26.690 --> 00:35:28.090 align:middle line:90% I showed you the plastic bottle. 00:35:28.090 --> 00:35:31.030 align:middle line:84% It has lots of residual stress in blow-molded bottle there. 00:35:31.030 --> 00:35:33.470 align:middle line:84% And those residual stress can eventually 00:35:33.470 --> 00:35:35.335 align:middle line:84% lead to cracking and damage of the material. 00:35:35.335 --> 00:35:37.710 align:middle line:84% So for pharmaceutical industry where they absolutely want 00:35:37.710 --> 00:35:40.170 align:middle line:84% everything to be perfect, you cannot do that. 00:35:40.170 --> 00:35:45.910 align:middle line:90% 00:35:45.910 --> 00:35:49.710 align:middle line:84% So for these kind of vials, the process to make it 00:35:49.710 --> 00:35:52.150 align:middle line:90% is so-called converting. 00:35:52.150 --> 00:35:56.070 align:middle line:84% Converting essentially means that you have a long glass tube. 00:35:56.070 --> 00:35:59.190 align:middle line:84% They start out by cutting it into sections. 00:35:59.190 --> 00:36:01.230 align:middle line:84% They then sealed one end and then 00:36:01.230 --> 00:36:03.270 align:middle line:90% shape it into the vial shape. 00:36:03.270 --> 00:36:07.750 align:middle line:84% So that process gives you less deformation and less 00:36:07.750 --> 00:36:10.590 align:middle line:84% residual stress compared to a blow-molding process. 00:36:10.590 --> 00:36:13.030 align:middle line:84% However, the conversion process involves high temperature 00:36:13.030 --> 00:36:13.530 align:middle line:90% treatment. 00:36:13.530 --> 00:36:14.905 align:middle line:84% You need to have high temperature 00:36:14.905 --> 00:36:16.670 align:middle line:84% to actually induce viscous deformation 00:36:16.670 --> 00:36:18.350 align:middle line:90% and seal it at one end. 00:36:18.350 --> 00:36:20.830 align:middle line:84% So during high-temperature treatment, 00:36:20.830 --> 00:36:25.110 align:middle line:84% boron can preferentially leach out from the glass 00:36:25.110 --> 00:36:29.270 align:middle line:84% and it actually vaporizes in the form of borosilicate. 00:36:29.270 --> 00:36:33.590 align:middle line:84% So this vaporized boron, some of it 00:36:33.590 --> 00:36:37.350 align:middle line:84% can get redeposited in parts of the glass container which 00:36:37.350 --> 00:36:39.790 align:middle line:84% has a lower temperature during the process. 00:36:39.790 --> 00:36:42.990 align:middle line:84% So this redeposited boron is then 00:36:42.990 --> 00:36:44.990 align:middle line:84% going to attach to the surface of container 00:36:44.990 --> 00:36:49.050 align:middle line:84% forming these islands or patches with chemical inhomogeneity. 00:36:49.050 --> 00:36:54.310 align:middle line:84% Then, the remaining glass will become boron-depleted. 00:36:54.310 --> 00:36:57.430 align:middle line:84% So boron-depleted portion becomes more 00:36:57.430 --> 00:36:59.710 align:middle line:90% susceptible to chemical attack. 00:36:59.710 --> 00:37:02.970 align:middle line:84% Remember, when glass comes into contact with aqueous solution, 00:37:02.970 --> 00:37:04.670 align:middle line:84% any kind of alkaline atom can leach 00:37:04.670 --> 00:37:06.087 align:middle line:84% reach out into the solution there, 00:37:06.087 --> 00:37:08.045 align:middle line:84% and it eventually increases the pH value, which 00:37:08.045 --> 00:37:09.390 align:middle line:90% accelerates a chemical attack. 00:37:09.390 --> 00:37:14.293 align:middle line:84% So then essentially these weaker part of the glasses, 00:37:14.293 --> 00:37:16.710 align:middle line:84% they will then delaminate from the glass container surface 00:37:16.710 --> 00:37:18.930 align:middle line:84% and essentially form small tiny little flakes 00:37:18.930 --> 00:37:21.150 align:middle line:90% that goes into the solution. 00:37:21.150 --> 00:37:26.670 align:middle line:84% So delamination turns out to be one of the biggest problems that 00:37:26.670 --> 00:37:32.350 align:middle line:84% compromised the yield of this kind of vaccine product. 00:37:32.350 --> 00:37:36.070 align:middle line:84% So that's why Corning want to go with this kind 00:37:36.070 --> 00:37:38.990 align:middle line:84% of boron-free formulation, which helps you to essentially 00:37:38.990 --> 00:37:42.830 align:middle line:84% minimize this kind of a boron volatilization issue 00:37:42.830 --> 00:37:46.710 align:middle line:84% and reduce this kind of delamination impact. 00:37:46.710 --> 00:37:48.250 align:middle line:90% So that's one thing. 00:37:48.250 --> 00:37:51.390 align:middle line:84% Second thing is that they also develop a proprietary surface 00:37:51.390 --> 00:37:54.030 align:middle line:84% coating, which helps you to prevent any kind of wear 00:37:54.030 --> 00:37:54.690 align:middle line:90% and tear. 00:37:54.690 --> 00:37:56.870 align:middle line:84% As I said, this is important for increasing the manufacturing 00:37:56.870 --> 00:37:57.570 align:middle line:90% throughput. 00:37:57.570 --> 00:38:03.030 align:middle line:84% And they also do ion exchange to enhance the mechanical strength. 00:38:03.030 --> 00:38:05.302 align:middle line:84% If you look at the ion exchange process they use, 00:38:05.302 --> 00:38:07.510 align:middle line:84% this is actually a picture taken from their factories 00:38:07.510 --> 00:38:10.390 align:middle line:84% where they're actually manufacturing these glass vials. 00:38:10.390 --> 00:38:14.950 align:middle line:84% You see these huge cassettes where they essentially put down 00:38:14.950 --> 00:38:17.550 align:middle line:84% these vials, essentially layer by layer, 00:38:17.550 --> 00:38:21.150 align:middle line:84% and then they put all of them into a huge molten salt chemical 00:38:21.150 --> 00:38:22.230 align:middle line:90% bath here. 00:38:22.230 --> 00:38:25.950 align:middle line:84% So this is a very highly parallel process, 00:38:25.950 --> 00:38:29.210 align:middle line:84% completely automated, that gives you extremely high throughput. 00:38:29.210 --> 00:38:32.430 align:middle line:84% So you can actually manufacture billions of these bottles 00:38:32.430 --> 00:38:33.690 align:middle line:90% during a short period of time. 00:38:33.690 --> 00:38:39.950 align:middle line:90% 00:38:39.950 --> 00:38:42.790 align:middle line:84% All right, so far we've been talking 00:38:42.790 --> 00:38:45.830 align:middle line:84% about different kinds of strengthening processes 00:38:45.830 --> 00:38:48.350 align:middle line:90% by inducing compressive stress. 00:38:48.350 --> 00:38:52.950 align:middle line:84% I want to mention that to get the optimal performance of glass 00:38:52.950 --> 00:38:56.990 align:middle line:84% materials, you can also try to combine it with other materials 00:38:56.990 --> 00:38:59.630 align:middle line:84% to create a composite that helps you 00:38:59.630 --> 00:39:02.850 align:middle line:84% to further boost their properties, reduce damage. 00:39:02.850 --> 00:39:06.510 align:middle line:84% The most simple form of this kind of composite materials 00:39:06.510 --> 00:39:08.150 align:middle line:90% is lamination. 00:39:08.150 --> 00:39:11.750 align:middle line:84% So in laminated glass, basically you 00:39:11.750 --> 00:39:14.430 align:middle line:84% have alternating layers between glass and polymers. 00:39:14.430 --> 00:39:16.670 align:middle line:84% You have a glass, put a polymer sheet, 00:39:16.670 --> 00:39:18.570 align:middle line:84% sandwich it between another layer of glass. 00:39:18.570 --> 00:39:23.350 align:middle line:84% You can repeat the process to get multilayer laminates. 00:39:23.350 --> 00:39:25.650 align:middle line:84% Now the advantage, there are twofold advantage here. 00:39:25.650 --> 00:39:27.830 align:middle line:84% One is that these soft polymers essentially 00:39:27.830 --> 00:39:30.350 align:middle line:84% helps you absorb the deformation and helps you spread it 00:39:30.350 --> 00:39:31.690 align:middle line:90% over the larger glass sheet. 00:39:31.690 --> 00:39:36.270 align:middle line:84% Also, even if you fracture one pane of glass, 00:39:36.270 --> 00:39:40.350 align:middle line:84% the polymer tends to hold the broken pieces 00:39:40.350 --> 00:39:43.890 align:middle line:84% together so that you don't have large shards falling down, 00:39:43.890 --> 00:39:46.510 align:middle line:84% which may potentially injure people. 00:39:46.510 --> 00:39:49.550 align:middle line:84% So that's why this kind of laminated glass, this 00:39:49.550 --> 00:39:51.350 align:middle line:84% is a standard glass material people use 00:39:51.350 --> 00:39:52.930 align:middle line:90% in architectural applications. 00:39:52.930 --> 00:39:56.790 align:middle line:84% So I'm actually giving you one example here. 00:39:56.790 --> 00:39:59.470 align:middle line:84% This is the Skywalk at Grand Canyon. 00:39:59.470 --> 00:40:02.060 align:middle line:90% Any of you have been there? 00:40:02.060 --> 00:40:03.260 align:middle line:90% It doesn't matter. 00:40:03.260 --> 00:40:05.620 align:middle line:84% Most of the building classes, they 00:40:05.620 --> 00:40:06.900 align:middle line:90% are actually laminated glass. 00:40:06.900 --> 00:40:11.260 align:middle line:84% But in the Skywalk, they have basically a glass bridge 00:40:11.260 --> 00:40:12.520 align:middle line:90% with glass bottom. 00:40:12.520 --> 00:40:14.500 align:middle line:84% You can actually walk on top of Grand Canyon 00:40:14.500 --> 00:40:17.000 align:middle line:84% and you can look down, take pictures, post them on Instagram 00:40:17.000 --> 00:40:17.900 align:middle line:90% or whatever. 00:40:17.900 --> 00:40:18.780 align:middle line:90% That's what it's for. 00:40:18.780 --> 00:40:20.918 align:middle line:90% So, not going to go there? 00:40:20.918 --> 00:40:21.460 align:middle line:90% AUDIENCE: No. 00:40:21.460 --> 00:40:22.322 align:middle line:90% [LAUGHTER] 00:40:22.322 --> 00:40:24.280 align:middle line:84% JUEJUN HU: Not going to pay to go there, right? 00:40:24.280 --> 00:40:26.557 align:middle line:90% AUDIENCE: No. 00:40:26.557 --> 00:40:27.140 align:middle line:90% JUEJUN HU: OK. 00:40:27.140 --> 00:40:28.500 align:middle line:90% So that's one example. 00:40:28.500 --> 00:40:31.500 align:middle line:84% Now, you can also make this laminated glass actually 00:40:31.500 --> 00:40:32.620 align:middle line:90% bulletproof. 00:40:32.620 --> 00:40:34.200 align:middle line:84% I will show you a little video here. 00:40:34.200 --> 00:40:35.340 align:middle line:90% [VIDEO PLAYBACK] 00:40:35.340 --> 00:40:40.380 align:middle line:84% - What you might not know is performance flattening. 00:40:40.380 --> 00:40:41.780 align:middle line:90% - It's not very targeted. 00:40:41.780 --> 00:40:44.300 align:middle line:90% 00:40:44.300 --> 00:40:47.380 align:middle line:84% - Meet Trent Kimball from San Antonio, Texas. 00:40:47.380 --> 00:40:51.700 align:middle line:84% He's the head of a firm that makes armored cars. 00:40:51.700 --> 00:40:54.360 align:middle line:84% - We'll take a common, everyday vehicle, 00:40:54.360 --> 00:40:57.820 align:middle line:84% take out the entire interior and put in our armored materials 00:40:57.820 --> 00:41:02.620 align:middle line:84% in the doors, the windows, the roof, the floor, the fuel tanks, 00:41:02.620 --> 00:41:04.980 align:middle line:84% and basically make it a cocoon of armor 00:41:04.980 --> 00:41:07.500 align:middle line:84% for the occupants inside the vehicle. 00:41:07.500 --> 00:41:09.960 align:middle line:84% These vehicles are going to politicians, businessmen, 00:41:09.960 --> 00:41:14.500 align:middle line:84% celebrities, and we've even offered them for the Pope. 00:41:14.500 --> 00:41:16.860 align:middle line:84% - One day, Trent got so tired of people asking him 00:41:16.860 --> 00:41:19.820 align:middle line:84% if the glass in his vehicles could really stop a bullet, 00:41:19.820 --> 00:41:22.220 align:middle line:90% he decided to prove it. 00:41:22.220 --> 00:41:23.800 align:middle line:90% - My name is Lawrence Caso. 00:41:23.800 --> 00:41:25.460 align:middle line:90% This is my AK-47. 00:41:25.460 --> 00:41:27.500 align:middle line:84% And today I'm gonna shoot my boss. 00:41:27.500 --> 00:41:28.960 align:middle line:90% Don't try this at home. 00:41:28.960 --> 00:41:33.789 align:middle line:90% 00:41:33.789 --> 00:41:37.156 align:middle line:90% [GUNSHOTS] 00:41:37.156 --> 00:41:41.705 align:middle line:90% 00:41:41.705 --> 00:41:43.580 align:middle line:84% - The only time I've had a little bit of fear 00:41:43.580 --> 00:41:46.100 align:middle line:84% enter my body was actually looking at the shooter, 00:41:46.100 --> 00:41:48.380 align:middle line:84% and the shooter's eyes got really big. 00:41:48.380 --> 00:41:51.700 align:middle line:84% So that's the only time I actually 00:41:51.700 --> 00:41:53.300 align:middle line:90% felt a little bit nervous. 00:41:53.300 --> 00:41:55.510 align:middle line:84% - Shooting at my boss is actually pretty interesting. 00:41:55.510 --> 00:41:57.760 align:middle line:84% I bet there's a lot of people who would like to do it, 00:41:57.760 --> 00:41:58.860 align:middle line:90% but I like my boss. 00:41:58.860 --> 00:42:01.620 align:middle line:84% He's pretty cool, so I'm glad he's OK. 00:42:01.620 --> 00:42:04.700 align:middle line:84% - Yes, Trent lived and uploaded this astonishing clip 00:42:04.700 --> 00:42:09.220 align:middle line:84% to the internet, where it has almost 6 million hits. 00:42:09.220 --> 00:42:14.100 align:middle line:84% But exactly how did that windshield save his skin? 00:42:14.100 --> 00:42:17.060 align:middle line:84% - The way you make glass absorb the energy from something like 00:42:17.060 --> 00:42:20.980 align:middle line:84% an AK-47 is that you have to make laminated glass. 00:42:20.980 --> 00:42:23.300 align:middle line:84% So you'll have a layer of glass, and then 00:42:23.300 --> 00:42:25.592 align:middle line:84% you'll have a layer of a material called polycarbonate, 00:42:25.592 --> 00:42:27.425 align:middle line:84% and then you'll have another layer of glass, 00:42:27.425 --> 00:42:29.040 align:middle line:84% and another layer of polycarbonate, 00:42:29.040 --> 00:42:31.780 align:middle line:84% and then so on and so on and so on to make a thick window. 00:42:31.780 --> 00:42:33.380 align:middle line:90% - This is a composite. 00:42:33.380 --> 00:42:36.740 align:middle line:84% And composites are used when you want the properties of two 00:42:36.740 --> 00:42:39.260 align:middle line:84% different materials, you want to put them together 00:42:39.260 --> 00:42:40.440 align:middle line:90% for different purposes. 00:42:40.440 --> 00:42:42.500 align:middle line:90% [GUNSHOT] 00:42:42.500 --> 00:42:44.780 align:middle line:84% If you were to just to have a thin piece of plastic, 00:42:44.780 --> 00:42:46.640 align:middle line:84% a bullet would easily penetrate that. 00:42:46.640 --> 00:42:48.260 align:middle line:84% If you just have a piece of glass, 00:42:48.260 --> 00:42:50.700 align:middle line:84% the bullet would cause it to completely shatter. 00:42:50.700 --> 00:42:53.740 align:middle line:84% You put those together and the glass 00:42:53.740 --> 00:42:58.300 align:middle line:84% disperses that force of that bullet over a larger area, 00:42:58.300 --> 00:43:02.500 align:middle line:84% and then the plastic can absorb that force. 00:43:02.500 --> 00:43:06.520 align:middle line:84% - Aside from the AK-47 the M16 and other military-type weapons, 00:43:06.520 --> 00:43:10.100 align:middle line:84% we can also stop the hand grenades and also right up 00:43:10.100 --> 00:43:12.900 align:middle line:90% to the 50-caliber round. 00:43:12.900 --> 00:43:16.080 align:middle line:84% - So thanks to the laminated glass, Trent dodged not one, 00:43:16.080 --> 00:43:19.620 align:middle line:84% but three bullets from a lethal semi-automatic rifle. 00:43:19.620 --> 00:43:22.280 align:middle line:84% The really scary part was still to come. 00:43:22.280 --> 00:43:23.900 align:middle line:90% - I did not tell my wife. 00:43:23.900 --> 00:43:27.200 align:middle line:84% She wasn't going to know until after. 00:43:27.200 --> 00:43:28.643 align:middle line:90% [GUNSHOT] 00:43:28.643 --> 00:43:30.060 align:middle line:84% Now she says we can't do it again. 00:43:30.060 --> 00:43:30.980 align:middle line:90% [END PLAYBACK] 00:43:30.980 --> 00:43:33.580 align:middle line:84% JUEJUN HU: OK, so now this is actually 00:43:33.580 --> 00:43:35.880 align:middle line:84% some scientific literature to prove that indeed, 00:43:35.880 --> 00:43:38.780 align:middle line:84% laminated glass actually helps increase 00:43:38.780 --> 00:43:40.400 align:middle line:90% the strength of glass products. 00:43:40.400 --> 00:43:43.620 align:middle line:84% So you can see from here, if you have a monolithic piece of glass 00:43:43.620 --> 00:43:46.140 align:middle line:84% versus glass with laminates and different thickness 00:43:46.140 --> 00:43:48.740 align:middle line:84% of the polymer layers, you can see an apparent increase 00:43:48.740 --> 00:43:50.520 align:middle line:90% of the fracture strength. 00:43:50.520 --> 00:43:53.240 align:middle line:84% So this is scientifically validated. 00:43:53.240 --> 00:43:57.130 align:middle line:84% You can actually help improve the strength. 00:43:57.130 --> 00:43:59.810 align:middle line:84% Now I just want to mention that you can actually take 00:43:59.810 --> 00:44:03.530 align:middle line:90% this design to another level. 00:44:03.530 --> 00:44:06.650 align:middle line:84% In addition to increasing the strength of glass materials, 00:44:06.650 --> 00:44:08.730 align:middle line:84% you can also use it to actually create 00:44:08.730 --> 00:44:11.130 align:middle line:84% highly flexible and bendable glass devices 00:44:11.130 --> 00:44:14.730 align:middle line:84% by using essentially this kind of a laminated or multi-layer 00:44:14.730 --> 00:44:15.930 align:middle line:90% design. 00:44:15.930 --> 00:44:18.203 align:middle line:84% So what I'm showing you here on the left-hand side, 00:44:18.203 --> 00:44:20.370 align:middle line:84% this is how you would actually make a foldable glass 00:44:20.370 --> 00:44:21.450 align:middle line:90% device, as I mentioned. 00:44:21.450 --> 00:44:23.410 align:middle line:90% I showed you on the board. 00:44:23.410 --> 00:44:26.690 align:middle line:90% Basically, you bend it. 00:44:26.690 --> 00:44:28.830 align:middle line:84% At the top, you have a tensile strength, bottom, 00:44:28.830 --> 00:44:29.950 align:middle line:84% the compressive strength, in the middle, 00:44:29.950 --> 00:44:31.867 align:middle line:84% you have a neutral plane or neutral axis where 00:44:31.867 --> 00:44:33.290 align:middle line:90% stress essentially vanishes. 00:44:33.290 --> 00:44:35.063 align:middle line:84% Now, this design is actually useful 00:44:35.063 --> 00:44:36.730 align:middle line:84% because people use it to create all type 00:44:36.730 --> 00:44:39.210 align:middle line:84% of flexible electronic or flexible photonic device 00:44:39.210 --> 00:44:40.010 align:middle line:90% designs. 00:44:40.010 --> 00:44:42.930 align:middle line:84% What they will do is they can put their device, which 00:44:42.930 --> 00:44:45.790 align:middle line:84% is very sensitive strength, you don't want to damage it. 00:44:45.790 --> 00:44:48.730 align:middle line:84% They will try to locate it right in the middle here 00:44:48.730 --> 00:44:51.092 align:middle line:84% at the neutral plane where the stress vanishes. 00:44:51.092 --> 00:44:53.550 align:middle line:84% So then regardless of how you actually bend your structure, 00:44:53.550 --> 00:44:54.925 align:middle line:84% your critical device is not going 00:44:54.925 --> 00:44:57.010 align:middle line:84% to get damaged because the actual stress on that 00:44:57.010 --> 00:44:58.970 align:middle line:90% is actually 0. 00:44:58.970 --> 00:45:00.930 align:middle line:84% Now, one downside about this design 00:45:00.930 --> 00:45:07.930 align:middle line:84% here is that your devices must be embedded in your substrate. 00:45:07.930 --> 00:45:10.010 align:middle line:84% This could be an issue, for example, 00:45:10.010 --> 00:45:13.890 align:middle line:84% if you have devices that actually dissipate lots of heat. 00:45:13.890 --> 00:45:15.970 align:middle line:84% Usually the flexible material around it 00:45:15.970 --> 00:45:17.650 align:middle line:84% is some kind of polymer, which is pretty 00:45:17.650 --> 00:45:19.410 align:middle line:90% terrible thermal conductor. 00:45:19.410 --> 00:45:20.930 align:middle line:84% And putting something that generates 00:45:20.930 --> 00:45:25.050 align:middle line:84% heat embedded in a poor thermal conductor is never a good idea. 00:45:25.050 --> 00:45:28.172 align:middle line:84% You also have some applications where you actually 00:45:28.172 --> 00:45:30.630 align:middle line:84% want your device to interact with surrounding environments. 00:45:30.630 --> 00:45:32.172 align:middle line:84% For example, for sensing applications 00:45:32.172 --> 00:45:33.730 align:middle line:84% like for biomedical applications, 00:45:33.730 --> 00:45:35.890 align:middle line:84% you want to have the device actually exposed on the surface 00:45:35.890 --> 00:45:38.010 align:middle line:84% or near the surface rather than embedded very deep 00:45:38.010 --> 00:45:40.010 align:middle line:90% in the substrate. 00:45:40.010 --> 00:45:43.690 align:middle line:84% So that's why an alternative design that was proposed here 00:45:43.690 --> 00:45:46.210 align:middle line:84% is to essentially create this kind of laminate structure. 00:45:46.210 --> 00:45:49.570 align:middle line:84% You have two stiff layers on top and bottom. 00:45:49.570 --> 00:45:53.110 align:middle line:84% In between, you put a very low Young's modulus material, 00:45:53.110 --> 00:45:55.850 align:middle line:90% sort of like rubber. 00:45:55.850 --> 00:45:58.352 align:middle line:84% Now when you bend it, what this new theory 00:45:58.352 --> 00:45:59.810 align:middle line:84% tells you is that instead of having 00:45:59.810 --> 00:46:02.490 align:middle line:84% just one single neutral axis, you actually 00:46:02.490 --> 00:46:04.930 align:middle line:84% have multiple planes of this kind of neutral axis 00:46:04.930 --> 00:46:07.730 align:middle line:84% here where the stress actually goes to 0. 00:46:07.730 --> 00:46:11.490 align:middle line:84% What's also more important is that by tuning the thickness 00:46:11.490 --> 00:46:14.050 align:middle line:84% and moduli of the different layers, 00:46:14.050 --> 00:46:17.330 align:middle line:84% you can tune the position of these neutral planes, 00:46:17.330 --> 00:46:21.130 align:middle line:84% essentially versatilely across the entire stack. 00:46:21.130 --> 00:46:24.370 align:middle line:84% So this allows you to put devices that's actually 00:46:24.370 --> 00:46:27.330 align:middle line:84% much closer to the surface compared 00:46:27.330 --> 00:46:29.980 align:middle line:84% to the classical neutral-plane design I'm 00:46:29.980 --> 00:46:31.230 align:middle line:90% showing on the left-hand side. 00:46:31.230 --> 00:46:36.770 align:middle line:90% 00:46:36.770 --> 00:46:41.410 align:middle line:84% So in this slide, I'm going to show you one demo. 00:46:41.410 --> 00:46:43.410 align:middle line:84% In this case, this is actually from my group. 00:46:43.410 --> 00:46:46.930 align:middle line:84% We were able to actually bend these flexible devices made out 00:46:46.930 --> 00:46:48.770 align:middle line:90% of glass materials. 00:46:48.770 --> 00:46:51.810 align:middle line:84% We can bend it down to an extremely small bending radius 00:46:51.810 --> 00:46:55.530 align:middle line:84% of 0.3 millimeters, and we can also essentially 00:46:55.530 --> 00:46:57.810 align:middle line:84% bend them for hundreds and thousands of cycles 00:46:57.810 --> 00:47:00.530 align:middle line:84% without seeing any damage, essentially leveraging this kind 00:47:00.530 --> 00:47:03.130 align:middle line:90% of a multi-neutral axis design. 00:47:03.130 --> 00:47:06.070 align:middle line:90% 00:47:06.070 --> 00:47:11.370 align:middle line:84% So this is for essentially to bend the glass. 00:47:11.370 --> 00:47:13.330 align:middle line:84% For some other applications, you not only want 00:47:13.330 --> 00:47:15.050 align:middle line:84% your material to be bendable, you also 00:47:15.050 --> 00:47:17.050 align:middle line:90% want it to be stretchable. 00:47:17.050 --> 00:47:19.950 align:middle line:84% So for example, for a lot of biomedical applications, 00:47:19.950 --> 00:47:21.990 align:middle line:84% people are trying to do epidermal integration, 00:47:21.990 --> 00:47:25.010 align:middle line:84% essentially putting the device directly on top of human skin. 00:47:25.010 --> 00:47:26.570 align:middle line:84% Human skin is actually a material 00:47:26.570 --> 00:47:29.870 align:middle line:84% that can easily withstand a large plastic deformation. 00:47:29.870 --> 00:47:32.930 align:middle line:84% So to match that, it means that your device should now 00:47:32.930 --> 00:47:33.910 align:middle line:90% be able to bend. 00:47:33.910 --> 00:47:36.650 align:middle line:84% It should be able to stretch and compress. 00:47:36.650 --> 00:47:40.810 align:middle line:84% So how do you create a stretchable device out 00:47:40.810 --> 00:47:45.010 align:middle line:90% of a stiff material? 00:47:45.010 --> 00:47:49.980 align:middle line:84% What you would do is now instead of doing multilayer lamination, 00:47:49.980 --> 00:47:53.575 align:middle line:84% you now change the in-plane configuration of the device. 00:47:53.575 --> 00:47:55.200 align:middle line:84% And the inspiration comes from the fact 00:47:55.200 --> 00:48:00.080 align:middle line:84% that, for example, if you think about a helix springs. 00:48:00.080 --> 00:48:02.860 align:middle line:84% Stainless steel is pretty stiff material in a bulk form, 00:48:02.860 --> 00:48:04.700 align:middle line:84% but if you shape it into this helical shape, 00:48:04.700 --> 00:48:06.560 align:middle line:84% you can actually stretch it and compress it 00:48:06.560 --> 00:48:08.280 align:middle line:90% with a very large strain. 00:48:08.280 --> 00:48:10.760 align:middle line:84% So what we did is you can now essentially 00:48:10.760 --> 00:48:15.560 align:middle line:84% create basically this kind of in-plane photonic springs. 00:48:15.560 --> 00:48:17.400 align:middle line:84% You essentially generate this kind 00:48:17.400 --> 00:48:21.280 align:middle line:84% of meandering shape of glass materials here. 00:48:21.280 --> 00:48:23.760 align:middle line:84% And when you actually stretch it, what we have shown 00:48:23.760 --> 00:48:26.240 align:middle line:84% is that it will deform essentially both in-plane 00:48:26.240 --> 00:48:27.640 align:middle line:90% and out of plane. 00:48:27.640 --> 00:48:30.840 align:middle line:84% And in doing so, the level of actual stress 00:48:30.840 --> 00:48:35.720 align:middle line:84% exerted on the glass material can be very, very small. 00:48:35.720 --> 00:48:40.440 align:middle line:84% On top of that, what we also do is that for some devices, 00:48:40.440 --> 00:48:42.880 align:middle line:84% you may have some critical devices where you absolutely 00:48:42.880 --> 00:48:45.720 align:middle line:84% don't want any kind of mechanical deformation. 00:48:45.720 --> 00:48:47.320 align:middle line:84% So what we can do is we can actually 00:48:47.320 --> 00:48:50.000 align:middle line:84% embed these critical devices that you don't want 00:48:50.000 --> 00:48:52.620 align:middle line:90% to deform in a stiff material. 00:48:52.620 --> 00:48:55.240 align:middle line:84% So eventually the configuration will 00:48:55.240 --> 00:48:57.840 align:middle line:84% be like you have these kind of stiff islands locally. 00:48:57.840 --> 00:49:00.000 align:middle line:84% They are interconnected together by these kind 00:49:00.000 --> 00:49:01.942 align:middle line:90% of meandering structures. 00:49:01.942 --> 00:49:04.400 align:middle line:84% So when you actually stretch it, essentially the meandering 00:49:04.400 --> 00:49:06.440 align:middle line:84% structure absorbs all the deformation 00:49:06.440 --> 00:49:10.000 align:middle line:84% while keeping the stiff islands completely intact, free 00:49:10.000 --> 00:49:11.040 align:middle line:90% of stress. 00:49:11.040 --> 00:49:14.140 align:middle line:84% So that's the strategy we have demonstrated. 00:49:14.140 --> 00:49:16.400 align:middle line:84% You can actually essentially create the structures out 00:49:16.400 --> 00:49:18.580 align:middle line:84% of essentially brittle glass material, 00:49:18.580 --> 00:49:21.420 align:middle line:84% but you can essentially create very long, large elongation, 00:49:21.420 --> 00:49:24.520 align:middle line:84% and you can also stretch them for thousands of cycles 00:49:24.520 --> 00:49:25.940 align:middle line:84% without seeing any kind of damage. 00:49:25.940 --> 00:49:29.435 align:middle line:90% 00:49:29.435 --> 00:49:31.810 align:middle line:84% So we've talked about two different kinds of strategies 00:49:31.810 --> 00:49:34.520 align:middle line:84% of how to change the configuration of glass materials 00:49:34.520 --> 00:49:36.760 align:middle line:84% to accommodate normal mechanical attributes. 00:49:36.760 --> 00:49:40.080 align:middle line:84% You can either do a outer plane structural engineering 00:49:40.080 --> 00:49:41.280 align:middle line:90% by lamination. 00:49:41.280 --> 00:49:43.320 align:middle line:84% You can also change the in-plane structure, 00:49:43.320 --> 00:49:48.280 align:middle line:84% like in the case of stretchable devices. 00:49:48.280 --> 00:49:50.320 align:middle line:90% You can combine the two. 00:49:50.320 --> 00:49:54.640 align:middle line:84% You can combine both lamination and also in-plane structuring. 00:49:54.640 --> 00:49:57.360 align:middle line:84% So this is shown in this example here. 00:49:57.360 --> 00:50:00.160 align:middle line:84% So in this paper, the authors essentially 00:50:00.160 --> 00:50:03.960 align:middle line:84% develop a toughened glass material inspired 00:50:03.960 --> 00:50:07.040 align:middle line:90% by this natural material nacre. 00:50:07.040 --> 00:50:13.000 align:middle line:84% So nacre is the material that is found inside the inner layer 00:50:13.000 --> 00:50:15.300 align:middle line:90% of the shells, seashells. 00:50:15.300 --> 00:50:18.680 align:middle line:84% It's essentially the lustrous material 00:50:18.680 --> 00:50:20.880 align:middle line:84% that people also call it the mother of pearl. 00:50:20.880 --> 00:50:27.840 align:middle line:84% So nacre, actually the main ingredient is calcium carbonate, 00:50:27.840 --> 00:50:30.160 align:middle line:90% limestone. 00:50:30.160 --> 00:50:33.660 align:middle line:84% But nacre, it turns out it's actually quite flexible 00:50:33.660 --> 00:50:35.000 align:middle line:90% and it's also very tough. 00:50:35.000 --> 00:50:38.240 align:middle line:84% So how do you make limestone essentially flexible and tough? 00:50:38.240 --> 00:50:42.240 align:middle line:84% So the trick here, if you look at the microstructure of nacre, 00:50:42.240 --> 00:50:44.400 align:middle line:90% it consists of multiple layers. 00:50:44.400 --> 00:50:48.800 align:middle line:84% And then within each layer, it's also not a monolithic piece. 00:50:48.800 --> 00:50:53.360 align:middle line:84% You have these kind of little platelets, essentially 00:50:53.360 --> 00:50:54.840 align:middle line:90% little plates. 00:50:54.840 --> 00:50:59.720 align:middle line:84% In between, you infiltrated with essentially a polymer material. 00:50:59.720 --> 00:51:03.320 align:middle line:84% So when you actually try to bend it or try to deform it, 00:51:03.320 --> 00:51:07.440 align:middle line:84% these platelets can essentially shift with respect to each other 00:51:07.440 --> 00:51:09.960 align:middle line:84% to absorb this kind of deformation there. 00:51:09.960 --> 00:51:14.920 align:middle line:84% So that's actually one of the configurational designs 00:51:14.920 --> 00:51:16.460 align:middle line:90% that the authors explored. 00:51:16.460 --> 00:51:18.600 align:middle line:84% So what they did is they started out essentially 00:51:18.600 --> 00:51:21.800 align:middle line:84% from just a typical flat sheet glass. 00:51:21.800 --> 00:51:24.880 align:middle line:84% They use laser writing to engrave these changes 00:51:24.880 --> 00:51:28.753 align:middle line:84% in these glass materials here without completely 00:51:28.753 --> 00:51:29.920 align:middle line:90% etching all the way through. 00:51:29.920 --> 00:51:32.000 align:middle line:84% So they are still connected as one piece. 00:51:32.000 --> 00:51:34.800 align:middle line:84% They then go through the typical lamination process 00:51:34.800 --> 00:51:36.640 align:middle line:90% to form a multilayer structure. 00:51:36.640 --> 00:51:39.200 align:middle line:84% And in the end, when you actually start to bend it, 00:51:39.200 --> 00:51:41.800 align:middle line:84% because the glass pieces, they are only connected together 00:51:41.800 --> 00:51:45.610 align:middle line:84% through this the deep trenches, so when you actually bend it, 00:51:45.610 --> 00:51:48.230 align:middle line:84% they essentially break and separate into individual pieces. 00:51:48.230 --> 00:51:49.910 align:middle line:84% So that's how they create this kind 00:51:49.910 --> 00:51:52.630 align:middle line:90% of nacre-like microstructures. 00:51:52.630 --> 00:51:55.270 align:middle line:84% They choose this kind of a glass material 00:51:55.270 --> 00:51:58.950 align:middle line:84% and the polymer material to have a very similar refractive index. 00:51:58.950 --> 00:52:00.560 align:middle line:84% So when the light goes through it, 00:52:00.560 --> 00:52:02.310 align:middle line:84% it barely sees this kind of microstructure 00:52:02.310 --> 00:52:03.550 align:middle line:84% there with minimal amount of scattering 00:52:03.550 --> 00:52:04.790 align:middle line:90% because of index matching. 00:52:04.790 --> 00:52:07.373 align:middle line:84% So that's why, if you look at the optical transmission diagram 00:52:07.373 --> 00:52:10.810 align:middle line:84% here, this kind of laminated material, 3D material, 00:52:10.810 --> 00:52:13.390 align:middle line:84% they still have optical transmittance that's fairly 00:52:13.390 --> 00:52:17.310 align:middle line:84% close or reasonably close compared to untreated glass. 00:52:17.310 --> 00:52:22.230 align:middle line:84% But they do exhibit much enhanced properties 00:52:22.230 --> 00:52:24.070 align:middle line:84% when it comes to resisting damage 00:52:24.070 --> 00:52:25.990 align:middle line:90% during mechanical deformation. 00:52:25.990 --> 00:52:31.230 align:middle line:84% So at the bottom here, you see that on top row, 00:52:31.230 --> 00:52:34.350 align:middle line:84% this is when you try to apply stress 00:52:34.350 --> 00:52:36.830 align:middle line:84% on one point on different kind of materials. 00:52:36.830 --> 00:52:40.750 align:middle line:84% You have a plain laminated glass, borosilicate and PMMA. 00:52:40.750 --> 00:52:43.050 align:middle line:84% At the bottom, you have this one. 00:52:43.050 --> 00:52:45.270 align:middle line:84% This is actually an nacre-like material. 00:52:45.270 --> 00:52:47.910 align:middle line:84% You clearly see that all the other cases, you generate 00:52:47.910 --> 00:52:49.650 align:middle line:90% lots of defects and cracks. 00:52:49.650 --> 00:52:52.430 align:middle line:84% In this case, the material remains intact. 00:52:52.430 --> 00:52:55.430 align:middle line:84% In terms of the mechanism, this can actually 00:52:55.430 --> 00:52:57.910 align:middle line:84% help absorb this kind of deformation here. 00:52:57.910 --> 00:52:59.330 align:middle line:90% I have this little video. 00:52:59.330 --> 00:53:01.990 align:middle line:84% And also I show the simulation result on the left-hand side. 00:53:01.990 --> 00:53:04.070 align:middle line:90% The bottom line is what I said. 00:53:04.070 --> 00:53:06.390 align:middle line:84% When you actually try to bend it, 00:53:06.390 --> 00:53:09.700 align:middle line:84% the platelets inside each laminate layers, 00:53:09.700 --> 00:53:11.950 align:middle line:84% they can essentially shift their position with respect 00:53:11.950 --> 00:53:12.850 align:middle line:90% to each other. 00:53:12.850 --> 00:53:15.410 align:middle line:84% And this helps to prevent the formation of cracks. 00:53:15.410 --> 00:53:17.345 align:middle line:84% You can clearly see from here this laminate 00:53:17.345 --> 00:53:18.970 align:middle line:84% has to move with respect to each other. 00:53:18.970 --> 00:53:21.110 align:middle line:84% This one is just a plain laminated sheet 00:53:21.110 --> 00:53:22.810 align:middle line:84% without this in-plane structuring. 00:53:22.810 --> 00:53:25.942 align:middle line:90% 00:53:25.942 --> 00:53:27.510 align:middle line:84% So this is one example to show you 00:53:27.510 --> 00:53:30.910 align:middle line:84% that by combining this actually out-of-plane and in-plane 00:53:30.910 --> 00:53:35.310 align:middle line:84% structuring, you can get to glass components that offer 00:53:35.310 --> 00:53:37.190 align:middle line:90% unique mechanical attributes. 00:53:37.190 --> 00:53:40.710 align:middle line:90% 00:53:40.710 --> 00:53:41.510 align:middle line:90% All right. 00:53:41.510 --> 00:53:43.470 align:middle line:84% I'm going to wrap up here for this section 00:53:43.470 --> 00:53:49.292 align:middle line:90% here on glass strengthening. 00:53:49.292 --> 00:53:51.750 align:middle line:84% The key takeaway message here is that we spend lots of time 00:53:51.750 --> 00:53:53.710 align:middle line:84% essentially talking about these different kind of strengthening 00:53:53.710 --> 00:53:54.510 align:middle line:90% techniques. 00:53:54.510 --> 00:53:56.030 align:middle line:84% So you want to be able to understand 00:53:56.030 --> 00:53:58.710 align:middle line:84% the differences between these strengthening methods. 00:53:58.710 --> 00:54:02.310 align:middle line:84% You have essentially a thermal tempering. 00:54:02.310 --> 00:54:05.870 align:middle line:84% For thermal tempering, there are two subcategories there. 00:54:05.870 --> 00:54:08.550 align:middle line:84% You can have heat-strengthened glass versus tempered glass. 00:54:08.550 --> 00:54:10.870 align:middle line:84% They really differ only by the amount of surface stress 00:54:10.870 --> 00:54:11.570 align:middle line:90% you accumulate. 00:54:11.570 --> 00:54:14.710 align:middle line:84% But for tempered glass, large surface stress 00:54:14.710 --> 00:54:17.550 align:middle line:84% also means that there's very large 00:54:17.550 --> 00:54:20.310 align:middle line:84% tensile stress region inside the glass material, which 00:54:20.310 --> 00:54:23.070 align:middle line:84% may cause spontaneous breaking, while the risk 00:54:23.070 --> 00:54:24.830 align:middle line:84% of spontaneous breaking is actually 00:54:24.830 --> 00:54:28.270 align:middle line:84% nonexistent for the case of heat-strengthened glass. 00:54:28.270 --> 00:54:30.330 align:middle line:84% For chemically strengthened glass material, 00:54:30.330 --> 00:54:34.150 align:middle line:84% they don't break spontaneously because of the smaller surface 00:54:34.150 --> 00:54:37.350 align:middle line:84% thickness of the chemical stress there. 00:54:37.350 --> 00:54:40.710 align:middle line:84% And one important difference I also want to highlight here 00:54:40.710 --> 00:54:43.230 align:middle line:84% is the temperature for the heat treatment. 00:54:43.230 --> 00:54:45.510 align:middle line:84% All of them involve treating the glass material 00:54:45.510 --> 00:54:46.590 align:middle line:90% at high temperature. 00:54:46.590 --> 00:54:49.630 align:middle line:84% But for heat strengthening, you need 00:54:49.630 --> 00:54:52.710 align:middle line:84% to heat it up to above the glass transition temperature. 00:54:52.710 --> 00:54:56.070 align:middle line:84% So a high temperature is actually stress free. 00:54:56.070 --> 00:54:58.430 align:middle line:84% You only accumulate this stress when you actually 00:54:58.430 --> 00:55:01.070 align:middle line:84% cool down such as the surface starts to approach the room 00:55:01.070 --> 00:55:01.870 align:middle line:90% temperature. 00:55:01.870 --> 00:55:04.110 align:middle line:84% That's when the temperature gradient is converted 00:55:04.110 --> 00:55:06.310 align:middle line:90% into the thermal stress. 00:55:06.310 --> 00:55:09.230 align:middle line:84% Versus chemical strengthening-- the process 00:55:09.230 --> 00:55:13.030 align:middle line:84% has to take place below the glass transition temperature, 00:55:13.030 --> 00:55:15.990 align:middle line:84% because you don't want any of the stress coming from the ion 00:55:15.990 --> 00:55:19.230 align:middle line:84% exchange process to get relaxed in this process. 00:55:19.230 --> 00:55:21.090 align:middle line:90% So that's a key difference here. 00:55:21.090 --> 00:55:22.482 align:middle line:84% The heat treatment temperature is 00:55:22.482 --> 00:55:24.190 align:middle line:84% different from these different processes. 00:55:24.190 --> 00:55:28.390 align:middle line:90% 00:55:28.390 --> 00:55:32.350 align:middle line:84% Any more questions about strengthening? 00:55:32.350 --> 00:55:33.350 align:middle line:90% OK. 00:55:33.350 --> 00:55:34.990 align:middle line:90% So we still have a bit of time. 00:55:34.990 --> 00:55:36.910 align:middle line:84% I think I actually have some time 00:55:36.910 --> 00:55:40.230 align:middle line:90% to quickly go over the P-set. 00:55:40.230 --> 00:55:44.370 align:middle line:84% OK, so first question is really straightforward. 00:55:44.370 --> 00:55:47.157 align:middle line:84% You all get it right one way or the other. 00:55:47.157 --> 00:55:48.990 align:middle line:84% I just want to mention that you can actually 00:55:48.990 --> 00:55:51.670 align:middle line:90% be a bit more quantitative. 00:55:51.670 --> 00:55:54.910 align:middle line:84% So co-working is not a way to actually make 00:55:54.910 --> 00:55:58.050 align:middle line:84% amorphous material pieces because of the strain softening. 00:55:58.050 --> 00:56:00.350 align:middle line:84% You can either do casting or you can do essentially 00:56:00.350 --> 00:56:01.610 align:middle line:90% a thermoplastic forming. 00:56:01.610 --> 00:56:03.910 align:middle line:84% So for casting, you can actually try 00:56:03.910 --> 00:56:06.910 align:middle line:84% to calculate what is the critical thickness you 00:56:06.910 --> 00:56:09.990 align:middle line:84% can cast without inducing crystallization. 00:56:09.990 --> 00:56:13.310 align:middle line:84% This data, this is what I found from online. 00:56:13.310 --> 00:56:15.470 align:middle line:84% You can also find the kinetics of crystallization 00:56:15.470 --> 00:56:17.950 align:middle line:84% by looking at the TT diagram for vitreloy. 00:56:17.950 --> 00:56:22.190 align:middle line:84% It turns out vitreloy is one of probably most widely used 00:56:22.190 --> 00:56:25.550 align:middle line:84% metallic glass compositions, so its data 00:56:25.550 --> 00:56:26.810 align:middle line:90% is pretty much everywhere. 00:56:26.810 --> 00:56:30.350 align:middle line:84% In fact, in my class when I show this TT diagram, 00:56:30.350 --> 00:56:32.750 align:middle line:90% the example I use is vitreloy. 00:56:32.750 --> 00:56:34.590 align:middle line:84% So you can actually get the critical cooling 00:56:34.590 --> 00:56:36.980 align:middle line:84% curve, critical cooling rate from this plot here. 00:56:36.980 --> 00:56:38.020 align:middle line:90% Plug it back. 00:56:38.020 --> 00:56:42.820 align:middle line:84% What you find out is that you can cast a vitreloy medallion up 00:56:42.820 --> 00:56:45.660 align:middle line:84% to 2.2 centimeters thickness without inducing 00:56:45.660 --> 00:56:46.360 align:middle line:90% crystallization. 00:56:46.360 --> 00:56:48.080 align:middle line:84% That's why it's called bulk metallic glasses. 00:56:48.080 --> 00:56:49.538 align:middle line:84% They really have quite good ability 00:56:49.538 --> 00:56:51.120 align:middle line:90% to resist crystallization. 00:56:51.120 --> 00:56:53.740 align:middle line:84% So you can treat them using conventional process. 00:56:53.740 --> 00:56:56.820 align:middle line:84% Now another route, you can do thermoplastic forming, 00:56:56.820 --> 00:56:58.398 align:middle line:84% for example, compressive molding. 00:56:58.398 --> 00:57:00.940 align:middle line:84% Compression molding should work pretty well because medallion 00:57:00.940 --> 00:57:02.120 align:middle line:90% is a pretty open structure. 00:57:02.120 --> 00:57:04.800 align:middle line:84% You can have two pieces, just press it and then it will mold. 00:57:04.800 --> 00:57:08.020 align:middle line:84% Again, you can do some quantitative calculation here. 00:57:08.020 --> 00:57:11.840 align:middle line:84% To predict the compressive molding process compatibility, 00:57:11.840 --> 00:57:15.100 align:middle line:84% you need to know how long the glass can stay at a state 00:57:15.100 --> 00:57:18.420 align:middle line:84% without crystallization during the process, 00:57:18.420 --> 00:57:19.720 align:middle line:90% at the right viscosity value. 00:57:19.720 --> 00:57:21.980 align:middle line:84% So you start out by essentially looking 00:57:21.980 --> 00:57:24.100 align:middle line:84% at the viscosity value for compression molding 00:57:24.100 --> 00:57:25.140 align:middle line:90% from the slides. 00:57:25.140 --> 00:57:31.120 align:middle line:84% Then you go to the Angel plot, which I also have in the slides. 00:57:31.120 --> 00:57:35.580 align:middle line:84% This one is vitreloy, so basically this slide here. 00:57:35.580 --> 00:57:37.340 align:middle line:84% So from that, you can then read out 00:57:37.340 --> 00:57:39.660 align:middle line:84% for this particular viscosity value what's 00:57:39.660 --> 00:57:41.420 align:middle line:90% the corresponding temperature. 00:57:41.420 --> 00:57:45.240 align:middle line:84% Then with that temperature, you then go back to TT diagram. 00:57:45.240 --> 00:57:46.980 align:middle line:84% You can then read out what's the time 00:57:46.980 --> 00:57:48.620 align:middle line:90% it takes for it to crystallize. 00:57:48.620 --> 00:57:50.378 align:middle line:84% Then that gives you the time window 00:57:50.378 --> 00:57:52.420 align:middle line:84% you can actually perform your compressed modeling 00:57:52.420 --> 00:57:54.240 align:middle line:84% without a risk of crystallization. 00:57:54.240 --> 00:57:56.722 align:middle line:90% So that's the calculation I did. 00:57:56.722 --> 00:57:58.180 align:middle line:84% We figured that we can actually get 00:57:58.180 --> 00:58:00.277 align:middle line:84% to a process window about more than three minutes, 00:58:00.277 --> 00:58:02.860 align:middle line:84% which should be more than enough for simple compressed molding 00:58:02.860 --> 00:58:03.480 align:middle line:90% process. 00:58:03.480 --> 00:58:06.220 align:middle line:84% So that's why we conclude this compressed molding 00:58:06.220 --> 00:58:09.580 align:middle line:84% process is compatible with structuring with vitreloy. 00:58:09.580 --> 00:58:13.740 align:middle line:84% You can also do injection molding. 00:58:13.740 --> 00:58:16.720 align:middle line:84% Injection molding requires lower viscosity, 00:58:16.720 --> 00:58:19.460 align:middle line:84% which then means that you have to go to higher temperature. 00:58:19.460 --> 00:58:23.620 align:middle line:84% In fact, as a result, you have a narrow time window. 00:58:23.620 --> 00:58:26.020 align:middle line:84% In this particular case, this is probably also not 00:58:26.020 --> 00:58:28.340 align:middle line:84% quite preferred because the processing temperature 00:58:28.340 --> 00:58:30.442 align:middle line:84% is actually above the nose temperature, which 00:58:30.442 --> 00:58:32.400 align:middle line:84% means that when you are trying to cool it down, 00:58:32.400 --> 00:58:35.025 align:middle line:84% you have to go through this nose temperature, which potentially 00:58:35.025 --> 00:58:37.200 align:middle line:84% has a much higher risk of crystallization. 00:58:37.200 --> 00:58:40.020 align:middle line:84% So it's not quite preferred but potentially also viable. 00:58:40.020 --> 00:58:43.080 align:middle line:90% So that's the first one. 00:58:43.080 --> 00:58:46.500 align:middle line:84% Second one, basically we write our code. 00:58:46.500 --> 00:58:50.960 align:middle line:84% One thing I noticed is that for some of you, 00:58:50.960 --> 00:58:54.140 align:middle line:84% your answer shows a little bump in the cooling curve. 00:58:54.140 --> 00:58:57.960 align:middle line:84% That should not happen because when the glass relaxes, 00:58:57.960 --> 00:59:00.180 align:middle line:84% it should always relax towards the quotation mark, 00:59:00.180 --> 00:59:02.625 align:middle line:84% "equilibrium state," which is a supercooled liquid state. 00:59:02.625 --> 00:59:04.000 align:middle line:84% For this particular example here, 00:59:04.000 --> 00:59:05.540 align:middle line:84% the supercooled liquid state has a volume smaller 00:59:05.540 --> 00:59:07.900 align:middle line:84% than the glassy state, so the volume should always 00:59:07.900 --> 00:59:10.320 align:middle line:90% decrease when it relaxes. 00:59:10.320 --> 00:59:11.680 align:middle line:90% So it should not go back up. 00:59:11.680 --> 00:59:12.957 align:middle line:90% Volume always goes down. 00:59:12.957 --> 00:59:14.040 align:middle line:90% So double check your code. 00:59:14.040 --> 00:59:16.562 align:middle line:84% Maybe probably it's just a sign error or something. 00:59:16.562 --> 00:59:19.020 align:middle line:84% I want to spend a little bit of time talking about this guy 00:59:19.020 --> 00:59:23.920 align:middle line:84% here, about the heat released in the heat treatment process. 00:59:23.920 --> 00:59:27.960 align:middle line:90% 00:59:27.960 --> 00:59:31.940 align:middle line:84% So heat released, this is directly related 00:59:31.940 --> 00:59:36.100 align:middle line:90% to enthalpy change. 00:59:36.100 --> 00:59:39.900 align:middle line:84% This is the amount of heat that's actually released. 00:59:39.900 --> 00:59:49.980 align:middle line:84% So this can be written as, the heat released, 00:59:49.980 --> 00:59:51.700 align:middle line:90% it should be this one. 00:59:51.700 --> 00:59:56.380 align:middle line:84% The heat flux, I should say, it should be the enthalpy change 00:59:56.380 --> 00:59:57.780 align:middle line:90% over time. 00:59:57.780 --> 01:00:03.580 align:middle line:84% So then this can be written, you can write it in this form here. 01:00:03.580 --> 01:00:10.380 align:middle line:84% Convert it into two derivatives here. 01:00:10.380 --> 01:00:14.320 align:middle line:84% The first one, dH/dT, this is actually quite straightforward. 01:00:14.320 --> 01:00:18.580 align:middle line:84% This basically is the slope of the enthalpy temperature curve. 01:00:18.580 --> 01:00:23.340 align:middle line:84% The second one is just your heating rate. 01:00:23.340 --> 01:00:25.980 align:middle line:84% It's a bit easier to try to plot the heat 01:00:25.980 --> 01:00:27.632 align:middle line:84% flux as a function of time, rather than 01:00:27.632 --> 01:00:28.840 align:middle line:90% as a function of temperature. 01:00:28.840 --> 01:00:30.610 align:middle line:84% It gets pretty confusing as a function of temperature 01:00:30.610 --> 01:00:32.568 align:middle line:84% because we are heating, cooling, and reheating. 01:00:32.568 --> 01:00:34.230 align:middle line:84% So the curve becomes a bit messy. 01:00:34.230 --> 01:00:36.370 align:middle line:84% I should be a bit more clear, I apologize. 01:00:36.370 --> 01:00:38.450 align:middle line:84% We should try to plot it as a function of time, 01:00:38.450 --> 01:00:40.158 align:middle line:84% rather than as a function of temperature. 01:00:40.158 --> 01:00:44.537 align:middle line:84% So I should have made it a bit less confusing. 01:00:44.537 --> 01:00:46.870 align:middle line:84% But anyway, if you try to plot it as a function of time, 01:00:46.870 --> 01:00:49.290 align:middle line:84% you need to look at the sign of these two factors here. 01:00:49.290 --> 01:00:51.490 align:middle line:84% So for this one, first term, we have 01:00:51.490 --> 01:00:54.610 align:middle line:84% mentioned that the shape of the curve for enthalpy 01:00:54.610 --> 01:00:56.730 align:middle line:84% is quite similar to that of volume. 01:00:56.730 --> 01:00:59.170 align:middle line:84% So we can basically go back to this plot 01:00:59.170 --> 01:01:01.130 align:middle line:84% here that shows how the volume changes 01:01:01.130 --> 01:01:03.170 align:middle line:84% as a function of temperature and essentially 01:01:03.170 --> 01:01:06.690 align:middle line:84% plot it, in this case, also for enthalpy. 01:01:06.690 --> 01:01:11.390 align:middle line:84% The caveat is that when I'm plotting this plot here, 01:01:11.390 --> 01:01:16.090 align:middle line:84% you notice that this last line is completely flat. 01:01:16.090 --> 01:01:17.990 align:middle line:84% Now, it's not supposed to be completely flat. 01:01:17.990 --> 01:01:19.407 align:middle line:84% The reason it's flat is because we 01:01:19.407 --> 01:01:22.250 align:middle line:84% neglect any kind of thermal expansion of the glassy state. 01:01:22.250 --> 01:01:25.550 align:middle line:90% Now, same thing for enthalpy. 01:01:25.550 --> 01:01:27.170 align:middle line:84% The slope of the line corresponds 01:01:27.170 --> 01:01:31.230 align:middle line:84% to the heat capacity of the glassy state, which is not zero. 01:01:31.230 --> 01:01:34.210 align:middle line:84% So we should essentially artificially add that slope 01:01:34.210 --> 01:01:34.730 align:middle line:90% back. 01:01:34.730 --> 01:01:37.030 align:middle line:84% So what we have will be something like this. 01:01:37.030 --> 01:01:38.570 align:middle line:90% You have this curve here. 01:01:38.570 --> 01:01:41.410 align:middle line:90% It goes back to the glass line. 01:01:41.410 --> 01:01:43.110 align:middle line:84% And then when you heat it back up, 01:01:43.110 --> 01:01:47.310 align:middle line:84% it will retrace this line here, goes back down, 01:01:47.310 --> 01:01:50.730 align:middle line:84% come here, and goes back, like that. 01:01:50.730 --> 01:01:53.770 align:middle line:84% This is what it should be looking like. 01:01:53.770 --> 01:01:57.290 align:middle line:84% So now we can divide it into a few different regimes 01:01:57.290 --> 01:01:59.470 align:middle line:84% and try to plot this diagram separately. 01:01:59.470 --> 01:02:05.510 align:middle line:84% So first regime is when we are cooling down. 01:02:05.510 --> 01:02:07.690 align:middle line:90% This is our regime 1. 01:02:07.690 --> 01:02:10.730 align:middle line:84% Second is when you start to reheat it up. 01:02:10.730 --> 01:02:12.970 align:middle line:84% Now during reheating process, you 01:02:12.970 --> 01:02:15.730 align:middle line:84% notice that there's actually a maximum point. 01:02:15.730 --> 01:02:19.690 align:middle line:84% So around this maximum point, the sign 01:02:19.690 --> 01:02:23.990 align:middle line:84% of the slope of the curve actually changes sign. 01:02:23.990 --> 01:02:25.910 align:middle line:84% So that's why we should have two regimes here. 01:02:25.910 --> 01:02:28.202 align:middle line:84% The first one is when you actually more or less retrace 01:02:28.202 --> 01:02:28.970 align:middle line:90% this curve. 01:02:28.970 --> 01:02:30.970 align:middle line:84% And then the third regime is when 01:02:30.970 --> 01:02:32.690 align:middle line:84% you see significant structural relaxation 01:02:32.690 --> 01:02:35.410 align:middle line:84% during the last transitional regime. 01:02:35.410 --> 01:02:37.590 align:middle line:84% And then the fourth regime, of course, 01:02:37.590 --> 01:02:41.130 align:middle line:84% is when you actually are cooling it back down. 01:02:41.130 --> 01:02:42.810 align:middle line:84% So now let's take a look at the sign 01:02:42.810 --> 01:02:46.950 align:middle line:84% of each of the terms during these four processes here. 01:02:46.950 --> 01:02:50.306 align:middle line:90% 01:02:50.306 --> 01:02:56.690 align:middle line:84% So first process, cooling, the slope of the curve is positive. 01:02:56.690 --> 01:03:03.530 align:middle line:90% This is regime 1. 01:03:03.530 --> 01:03:05.750 align:middle line:84% So the slope of a curve is positive. 01:03:05.750 --> 01:03:08.250 align:middle line:84% dT/dt, this is a cooling process, so 01:03:08.250 --> 01:03:11.650 align:middle line:84% temperature decreases with time, so this should be negative. 01:03:11.650 --> 01:03:18.930 align:middle line:84% So overall, the dH/dt for the heat flux, this 01:03:18.930 --> 01:03:20.950 align:middle line:90% should be a negative. 01:03:20.950 --> 01:03:25.930 align:middle line:90% 01:03:25.930 --> 01:03:27.450 align:middle line:84% Now it should be negative because 01:03:27.450 --> 01:03:31.850 align:middle line:84% during cooling the material obviously releases heat 01:03:31.850 --> 01:03:32.890 align:middle line:90% to the environment. 01:03:32.890 --> 01:03:34.440 align:middle line:84% So negative sign essentially just 01:03:34.440 --> 01:03:36.190 align:middle line:84% means that the material is releasing heat. 01:03:36.190 --> 01:03:38.890 align:middle line:90% 01:03:38.890 --> 01:03:44.650 align:middle line:84% Second regime, so for the second regime here, 01:03:44.650 --> 01:03:48.010 align:middle line:90% dH/dt is still positive. 01:03:48.010 --> 01:03:54.770 align:middle line:90% 01:03:54.770 --> 01:04:02.010 align:middle line:84% But dT/dt, this is the reheating process, so it becomes positive. 01:04:02.010 --> 01:04:05.010 align:middle line:84% So then what we have, this is a positive 0. 01:04:05.010 --> 01:04:07.010 align:middle line:84% It means that the material is now absorbing heat 01:04:07.010 --> 01:04:10.010 align:middle line:90% because you're reheating it up. 01:04:10.010 --> 01:04:15.250 align:middle line:84% OK, regime 3, for regime 3 what's different 01:04:15.250 --> 01:04:21.390 align:middle line:84% here is that this curve, dH/dt, now has a negative slope. 01:04:21.390 --> 01:04:23.080 align:middle line:90% This is the reheating process. 01:04:23.080 --> 01:04:28.340 align:middle line:84% So while the temperature is increasing, 01:04:28.340 --> 01:04:31.360 align:middle line:84% you're actually releasing heat because you essentially 01:04:31.360 --> 01:04:32.722 align:middle line:90% the glass material is relaxing. 01:04:32.722 --> 01:04:35.180 align:middle line:84% It's relaxing to this more stable supercooled liquid state. 01:04:35.180 --> 01:04:37.480 align:middle line:84% In the process, it's releasing excess enthalpy 01:04:37.480 --> 01:04:38.760 align:middle line:90% to the environment. 01:04:38.760 --> 01:04:41.880 align:middle line:90% That's why it becomes negative. 01:04:41.880 --> 01:04:45.880 align:middle line:84% Finally, the last process here, process number 4, 01:04:45.880 --> 01:04:50.380 align:middle line:84% again, you have a positive slope for dH/dT. dT/dt, 01:04:50.380 --> 01:04:54.400 align:middle line:84% this is cooling, so it should be negative. 01:04:54.400 --> 01:04:58.680 align:middle line:90% So it should also be negative. 01:04:58.680 --> 01:05:01.360 align:middle line:84% Now let's try to stitch things together and try 01:05:01.360 --> 01:05:05.520 align:middle line:90% to plot this diagram here. 01:05:05.520 --> 01:05:09.040 align:middle line:90% So this will be our dH/dt. 01:05:09.040 --> 01:05:12.177 align:middle line:84% This will be our heat flux basically as a function of time. 01:05:12.177 --> 01:05:13.260 align:middle line:90% So let's see what happens. 01:05:13.260 --> 01:05:18.480 align:middle line:84% So first part, this is going to be a negative, 01:05:18.480 --> 01:05:21.960 align:middle line:84% but initially, you get a larger slope. 01:05:21.960 --> 01:05:27.680 align:middle line:84% So for constant cooling rate, initially you 01:05:27.680 --> 01:05:33.640 align:middle line:90% get a larger negative heat flux. 01:05:33.640 --> 01:05:39.290 align:middle line:84% And then you should smoothly reduce to a lower level 01:05:39.290 --> 01:05:41.040 align:middle line:84% because now you are essentially converting 01:05:41.040 --> 01:05:43.860 align:middle line:84% from supercooled liquid to glass. 01:05:43.860 --> 01:05:45.620 align:middle line:84% And glass has a smaller heat capacity. 01:05:45.620 --> 01:05:49.500 align:middle line:90% 01:05:49.500 --> 01:05:51.720 align:middle line:90% Now, go to regime 2. 01:05:51.720 --> 01:05:54.380 align:middle line:84% I should probably make it a bit narrower. 01:05:54.380 --> 01:05:56.100 align:middle line:84% I don't have that much space on board. 01:05:56.100 --> 01:05:59.580 align:middle line:90% 01:05:59.580 --> 01:06:05.760 align:middle line:90% So this is regime 1, regime 1. 01:06:05.760 --> 01:06:13.880 align:middle line:84% Now regime 2, in regime 2, we have a positive. 01:06:13.880 --> 01:06:18.400 align:middle line:84% Initially, regime 2, the curve actually more or less retraces 01:06:18.400 --> 01:06:19.660 align:middle line:90% the curve in regime 1. 01:06:19.660 --> 01:06:23.040 align:middle line:84% So it means that the heat flux essentially just flips signs, 01:06:23.040 --> 01:06:26.640 align:middle line:90% but the magnitude is the same. 01:06:26.640 --> 01:06:30.760 align:middle line:84% And then eventually, because the dH/dt starts to decrease 01:06:30.760 --> 01:06:33.360 align:middle line:84% and eventually flips sign, it means 01:06:33.360 --> 01:06:36.660 align:middle line:84% that the heat flux should also decrease 01:06:36.660 --> 01:06:39.800 align:middle line:90% and eventually flips signs. 01:06:39.800 --> 01:06:41.500 align:middle line:90% So this would be regime 2. 01:06:41.500 --> 01:06:48.166 align:middle line:90% 01:06:48.166 --> 01:06:52.200 align:middle line:84% So that means regime 3, regime 3 and 2, 01:06:52.200 --> 01:06:53.500 align:middle line:90% they should be continuous. 01:06:53.500 --> 01:06:56.960 align:middle line:90% It's basically one same process. 01:06:56.960 --> 01:06:58.960 align:middle line:84% So the curve should also be continuous. 01:06:58.960 --> 01:07:01.920 align:middle line:90% It should continue to extend. 01:07:01.920 --> 01:07:06.760 align:middle line:84% And eventually it should reach some kind of maximum slope here. 01:07:06.760 --> 01:07:09.800 align:middle line:84% And then the slope would then decrease again. 01:07:09.800 --> 01:07:13.920 align:middle line:84% So it means that it should reach a maximum and then goes back. 01:07:13.920 --> 01:07:20.920 align:middle line:90% 01:07:20.920 --> 01:07:22.580 align:middle line:90% So this will be regime 3. 01:07:22.580 --> 01:07:34.160 align:middle line:90% 01:07:34.160 --> 01:07:46.080 align:middle line:84% Finally, for regime 4, regime 4, this would also be a negative, 01:07:46.080 --> 01:07:47.340 align:middle line:90% started out negative. 01:07:47.340 --> 01:07:50.640 align:middle line:90% 01:07:50.640 --> 01:07:55.460 align:middle line:84% And then the slope increases and then decreases again. 01:07:55.460 --> 01:08:00.360 align:middle line:90% 01:08:00.360 --> 01:08:01.020 align:middle line:90% Oh, sorry. 01:08:01.020 --> 01:08:03.440 align:middle line:84% I should make it a bit more dramatic. 01:08:03.440 --> 01:08:05.920 align:middle line:84% In the end, the slope is the same as the regime 01:08:05.920 --> 01:08:07.100 align:middle line:90% 2 and regime 1. 01:08:07.100 --> 01:08:10.080 align:middle line:84% So it should come back to the same level. 01:08:10.080 --> 01:08:12.320 align:middle line:84% So it should go larger and then eventually comes back 01:08:12.320 --> 01:08:14.640 align:middle line:90% to the same level. 01:08:14.640 --> 01:08:17.430 align:middle line:84% It should come back to the same level as the final plateau 01:08:17.430 --> 01:08:18.290 align:middle line:90% of regime 1. 01:08:18.290 --> 01:08:23.910 align:middle line:90% 01:08:23.910 --> 01:08:27.229 align:middle line:84% So that's what the curve is going to look like. 01:08:27.229 --> 01:08:31.590 align:middle line:84% This is the more complex problem in the whole P-set here. 01:08:31.590 --> 01:08:33.630 align:middle line:84% AUDIENCE: I was only a little bit wrong. 01:08:33.630 --> 01:08:35.672 align:middle line:84% JUEJUN HU: I think you more or less get it right. 01:08:35.672 --> 01:08:36.550 align:middle line:90% You get the idea. 01:08:36.550 --> 01:08:39.187 align:middle line:84% Part of your confusion, I should really 01:08:39.187 --> 01:08:41.270 align:middle line:84% say that you should plot it as a function of time, 01:08:41.270 --> 01:08:42.280 align:middle line:84% not as a function of temperature. 01:08:42.280 --> 01:08:43.370 align:middle line:84% If you plot it as function of temperature, 01:08:43.370 --> 01:08:45.439 align:middle line:84% it gets really confusing which way it's going 01:08:45.439 --> 01:08:47.590 align:middle line:90% and all that stuff. 01:08:47.590 --> 01:08:48.670 align:middle line:90% All right. 01:08:48.670 --> 01:08:52.856 align:middle line:84% Last question here, this one is straightforward. 01:08:52.856 --> 01:08:55.189 align:middle line:84% Basically, you just want to prevent the formation of ice 01:08:55.189 --> 01:08:57.478 align:middle line:90% crystals by amortizing it. 01:08:57.478 --> 01:08:59.270 align:middle line:84% The interesting part is that I was actually 01:08:59.270 --> 01:09:03.390 align:middle line:84% reading about this technology called cryonics. 01:09:03.390 --> 01:09:04.710 align:middle line:90% Maybe you all heard of it. 01:09:04.710 --> 01:09:08.189 align:middle line:84% It's also called, I think, cryogenic sleep 01:09:08.189 --> 01:09:10.550 align:middle line:84% or cryogenic hibernation, which appears 01:09:10.550 --> 01:09:13.590 align:middle line:90% in lots of science fiction. 01:09:13.590 --> 01:09:16.790 align:middle line:84% Basically, it's saying if you want to go into a hibernation 01:09:16.790 --> 01:09:18.710 align:middle line:84% state, if you can quickly freeze it, 01:09:18.710 --> 01:09:22.670 align:middle line:84% then you can wake back up after a few decades, 01:09:22.670 --> 01:09:24.250 align:middle line:90% as long as you heat up. 01:09:24.250 --> 01:09:27.229 align:middle line:90% 01:09:27.229 --> 01:09:30.910 align:middle line:84% Now, the biggest barrier why that doesn't work 01:09:30.910 --> 01:09:33.270 align:middle line:90% is because of crystallization. 01:09:33.270 --> 01:09:37.390 align:middle line:84% So when you actually try to freeze body fluids, 01:09:37.390 --> 01:09:39.513 align:middle line:84% essentially it forms ice crystals. 01:09:39.513 --> 01:09:41.430 align:middle line:84% And these tiny little sharp ice crystals, they 01:09:41.430 --> 01:09:44.310 align:middle line:84% can pierce through the cell membrane and then eventually 01:09:44.310 --> 01:09:46.390 align:middle line:84% leading to it becoming nonfunctional. 01:09:46.390 --> 01:09:51.189 align:middle line:84% So the whole idea for people working in that field 01:09:51.189 --> 01:09:53.790 align:middle line:84% is to create some material, some liquid, 01:09:53.790 --> 01:09:56.590 align:middle line:84% that can go into a glassy state as you 01:09:56.590 --> 01:09:58.230 align:middle line:90% go to cryogenic temperature. 01:09:58.230 --> 01:10:00.110 align:middle line:84% So when you go back up, you prevent any kind 01:10:00.110 --> 01:10:02.270 align:middle line:90% of formation of crystals. 01:10:02.270 --> 01:10:04.550 align:middle line:84% So amorphization, amorphous material 01:10:04.550 --> 01:10:07.830 align:middle line:84% is quite important there, if you want to realize essentially 01:10:07.830 --> 01:10:10.593 align:middle line:90% eternal life through cryonics. 01:10:10.593 --> 01:10:12.510 align:middle line:84% So that's why amorphous material is important. 01:10:12.510 --> 01:10:13.530 align:middle line:90% It's eternal. 01:10:13.530 --> 01:10:14.290 align:middle line:90% We have seen that. 01:10:14.290 --> 01:10:19.125 align:middle line:90% 01:10:19.125 --> 01:10:20.250 align:middle line:90% We still have five minutes. 01:10:20.250 --> 01:10:22.590 align:middle line:84% I think I can actually do the demo. 01:10:22.590 --> 01:10:25.730 align:middle line:84% Now, this lecture is about glass strengthening. 01:10:25.730 --> 01:10:28.310 align:middle line:84% Next lecture is actually about amorphous, 01:10:28.310 --> 01:10:32.070 align:middle line:90% about metallic glasses. 01:10:32.070 --> 01:10:34.390 align:middle line:84% I'm not going to go into details there because we 01:10:34.390 --> 01:10:38.350 align:middle line:84% don't have enough time, but I can show you a little demo. 01:10:38.350 --> 01:10:43.130 align:middle line:84% So metallic glasses has a lot of unique mechanical attributes. 01:10:43.130 --> 01:10:47.310 align:middle line:84% That's why I have a chapter dedicated to metallic glasses. 01:10:47.310 --> 01:10:53.090 align:middle line:84% So what I have here, on top this is a sheet of metallic glass, 01:10:53.090 --> 01:10:54.630 align:middle line:90% vitreloy. 01:10:54.630 --> 01:10:58.550 align:middle line:84% It's sitting on an aluminum truck. 01:10:58.550 --> 01:11:02.910 align:middle line:84% I have with me a little bowl made of stainless steel. 01:11:02.910 --> 01:11:05.750 align:middle line:84% What I will do, I will let it bounce back, initially 01:11:05.750 --> 01:11:08.350 align:middle line:84% starting on, let it bounce back on the aluminum. 01:11:08.350 --> 01:11:12.630 align:middle line:90% 01:11:12.630 --> 01:11:16.270 align:middle line:84% OK, I'll do it again, pretty boring. 01:11:16.270 --> 01:11:20.978 align:middle line:84% OK, now I turn to the face where I have metallic glass. 01:11:20.978 --> 01:11:22.270 align:middle line:90% I'll do exactly the same thing. 01:11:22.270 --> 01:11:31.770 align:middle line:90% 01:11:31.770 --> 01:11:32.930 align:middle line:90% I'll try again. 01:11:32.930 --> 01:11:43.750 align:middle line:90% 01:11:43.750 --> 01:11:45.190 align:middle line:90% You can do aluminum again. 01:11:45.190 --> 01:11:46.850 align:middle line:84% I don't know if it was properly filmed. 01:11:46.850 --> 01:11:53.150 align:middle line:90% 01:11:53.150 --> 01:11:55.990 align:middle line:90% Why do you see the difference? 01:11:55.990 --> 01:12:01.670 align:middle line:84% AUDIENCE: Because the glasses have really high modulus, 01:12:01.670 --> 01:12:03.430 align:middle line:90% elastic modulus. 01:12:03.430 --> 01:12:05.110 align:middle line:84% JUEJUN HU: What limits the height 01:12:05.110 --> 01:12:06.890 align:middle line:90% of the ball bouncing back? 01:12:06.890 --> 01:12:09.830 align:middle line:90% 01:12:09.830 --> 01:12:11.110 align:middle line:90% AUDIENCE: Plastic deformation. 01:12:11.110 --> 01:12:12.100 align:middle line:90% JUEJUN HU: OK. 01:12:12.100 --> 01:12:13.820 align:middle line:84% What does plastic deformation do? 01:12:13.820 --> 01:12:15.380 align:middle line:90% AUDIENCE: It absorbs energy. 01:12:15.380 --> 01:12:16.240 align:middle line:90% JUEJUN HU: Exactly. 01:12:16.240 --> 01:12:18.620 align:middle line:84% If it absorbs energy, then it doesn't bounce back 01:12:18.620 --> 01:12:21.220 align:middle line:90% nearly as high. 01:12:21.220 --> 01:12:27.020 align:middle line:84% The basic physics behind it, it's 01:12:27.020 --> 01:12:30.777 align:middle line:84% because aluminum, when actually the ball impinge on the surface, 01:12:30.777 --> 01:12:32.360 align:middle line:84% it causes lots of plastic deformation. 01:12:32.360 --> 01:12:33.580 align:middle line:90% That's why it causes dents. 01:12:33.580 --> 01:12:35.140 align:middle line:90% It absorbs energy. 01:12:35.140 --> 01:12:38.780 align:middle line:84% But for metallic glass, it doesn't have a proper mechanism 01:12:38.780 --> 01:12:40.980 align:middle line:84% to induce large plastic deformation. 01:12:40.980 --> 01:12:43.420 align:middle line:84% We mentioned about this because it doesn't 01:12:43.420 --> 01:12:45.540 align:middle line:90% have any kind of dislocations. 01:12:45.540 --> 01:12:47.620 align:middle line:84% So you don't have the typical dislocation motion 01:12:47.620 --> 01:12:49.960 align:middle line:84% mechanism to actually cause plastic deformation. 01:12:49.960 --> 01:12:51.660 align:middle line:84% And that's why metallic glasses tend 01:12:51.660 --> 01:12:54.220 align:middle line:84% to have much higher elastic limit and higher yield strength. 01:12:54.220 --> 01:12:56.175 align:middle line:84% So this experiment actually shows 01:12:56.175 --> 01:12:58.300 align:middle line:84% what happens when you actually have a material that 01:12:58.300 --> 01:12:59.720 align:middle line:90% doesn't deform plastically. 01:12:59.720 --> 01:13:00.900 align:middle line:84% Essentially, all the energy gets stored 01:13:00.900 --> 01:13:02.400 align:middle line:84% in the form of elastic energy, so it 01:13:02.400 --> 01:13:07.020 align:middle line:84% will bounce back with minimal amount of internal damping. 01:13:07.020 --> 01:13:08.240 align:middle line:90% I'm going to stop here. 01:13:08.240 --> 01:13:09.460 align:middle line:84% And in the next lecture, we are going 01:13:09.460 --> 01:13:11.710 align:middle line:84% through more details about the actual structural model 01:13:11.710 --> 01:13:14.260 align:middle line:84% of amorphous metal and its mechanical attributes. 01:13:14.260 --> 01:13:16.290 align:middle line:90% All right, thank you. 01:13:16.290 --> 01:14:09.000 align:middle line:90%